INDIAN JOURNAL OF TUBERCULOSIS
Official organ of the
Tuberculosis Association of India
Editor: Dr. Vol. XVI: No. 2 April 1969
Editorial : Planning, Conduct and Evaluation of
Dr. M.D. Deshmukh
Dr. N.L. Bordia Research Studies ... 33
Award of T.A.I. Gold Medal ... 35
D . H.B. Dingley Planning of Research Studies
Dr. S.P. Pamra —S.S. Nair ... 37
Planning, Conduct and Evaluation of Controlled
—S. Radhakrishna …42
Planning and Conduct of Epidemiological Surveys
—G.S. Acharyulu ... 47
Analysis and Presentation of Data
— G.P. Mathur ... 51
Agar Electrophoresis of Serum Proteins in Pulmo-
—K.L. Agrawal, S. Narasimha Rao and D.P. Agrawal ... 54
Ethionamide and Isoniazid in the Middle-aged
and the Elderly Patients
—R.K. Narang ... 60
Published quarterly in A Comparative Clinical Evaluation of the Role of
the months of January, Thiaocetazone and PAS in the Management of
April, July and October. Pulmonary Tuberculosis
—B.K. Khanna ... 64
Annual Subscription :
Rs. 15/-. £/21/-, $ 3.
Rs.4.00 News & Notes ** Astracts
* * *
Published on behalf of the Tuberculosis Association of India, by
the Secretary-General, 3, Red Cross Road, New Delhi-1
Procedures For Publication Of
The Sub-Committee appointed by the Technical
Committee of the Association have Lald down
following procedures regarding publication of
papers presented at the National Conference on
TB and Chest Diseases :—
(1) Papers approved by the Technical Com
mittee for presentation at the National Conference
be published in the Proceedings of the Conference.
(2) Papers presented at the National Con
ference can be reproduced in full or part thereof in
the Indian Journal of Tuberculosis and the author
intimated of the same.
(3) Papers presented at the National Con
ference may be reproduced in full or part thereof
in any other Journal after obtaining permission
from the Secretary-General of the Association and
the acknowledgement to be read as : “This paper
was presented at the .. . National Conference on
TB and Chest Diseases held at. . . and has been/is
being published in the Proceedings of the
Indian Journal of Tuberculosis
Vol. XVI New Delhi, April 1969 No. 2
PLANNING, CONDUCT AND EVALUATION OF RESEARCH STUDIES
An unusual observation and deductions therefrom for possible cause
of such an observation is, generally, the mother of research. A real research
worker will not rest content with this only. He will, thereafter, take suitable
steps to solve problems arising out of such an observation. This is how man
has delved into the depths of the unknown and collected figures and facts for
the benefit of mankind.
Many epoch-making discoveries were made in the past by individuals.
Because of rapid advances in medical and sister sciences, and development of
specialities, individual effort is being replaced by that of a “team” in solving
most of the problems. It has, however, to be conceded that the fountain of
research is still the individual with a bent of mind for careful observation
and deductions therefrom. To get the right answer or, in other words, to
fulfil the objective of the study, the planning, conduct and evaluation of the
findings should be entrusted to a suitable ‘team’ of experts.
Most often, specially for operational research, inclusion of a “Statis-
tician” is of vital importance in a research team.
It is obvious that such a “team” must work like a well-oiled machine.
For this purpose, all members of the team must clearly understand each
other’s views and stand-points. The problem to be investigated must be
defined, expressed in an unambiguous language and should, preferably, remain
clear of other issues. Terminologies to be used for accord must also be
defined in clear and simple language. Planning of the study should take into
account all possible sources of errors and biases with an effort to eliminate
or diminish them. Many other factors may also have to be considered. The
design of a study must, therefore, be built with great care in which every
participant should have his say and share of responsibilities. All these spokes
in the wheel of study must be in the realm of practicability, and should
strengthen confidence in the final interpretations and conclusions.
The practice of seeking help of a “Statistician” at a late stage for
interpretation of data and drawing conclusions is utterly wrong. The data
Ind, J. Tub., Vol. XVI, No. 2
can be correctly interpreted only when the design of a study is adequate
and information rightly recorded. Besides, and this is most important, the
data must be interpreted with common sense in relation to the purpose and
picture of the structure of the study. The Statistician must, therefore, be
associated from the start to the finish of a study. He should be deeply invol-
ved in planning, check its conduct, agree to any interim modification of the
plan necessitated by unforeseen conditions and finally have the largest say in
the evaluation. This is true both for a planned and retrospective study.
India is largely involved in the field of operational research to develop
a suitable tuberculosis control programme. Few well-equipped institutions
may have the resources to plan such studies. This is hardly possible for
others. At the same time, the subject of tuberculosis is such that data, res-
tricted by uneven distribution of regions, may have many failings. To obviate
this, well-planned and supervised co-operative studies, designed by a team of
experts, appears to be the best method. It increases immensely the confidence
not only in the conclusions but also their acceptability for the country’s cont-
Writing a paper on a study is also an art. The manner of presenta-
tion should differ in long and short publications and at Conferences. For the
latter, tables etc. to be projected should not usually contain details but should
be clean and highlight the important points only. In short, the presentation
should take into account the consumers and allotted space or time.
The four papers on Statistics published in this issue of the Journal
provide valuable guide-line in regard to utilisation of statistical methods in
Ind. J. Tub., Vol. XVI, No. 2
AWARD OF T.A.I. GOLD MEDAL
The Fourth Award of the Tuberculosis Association of India’s Gold
Medal for outstanding work in the Tuberculosis field was conferred on
Dr. P. V. Benjamin at the time of the 24th National Conference on Tuber-
culosis and Chest Diseases held in Trivandrum in January 1969.
tion of BCG Vaccination in India. He persua-
ded the various Universities to institute
Diploma Courses in Tuberculosis Diseases.
He set up a separate Section for Tuberculosis
in the Directorate General of Health Services.
He advocated domicilliary treatment of
patients and prepared suitable schemes for the
purpose in the national development pro-
gramme. He was responsible for the esta-
blishment of the National Tuberculosis
Institute in Bangalore, the Chemotherapy
Centre in Madras, the Tuberculosis Research
Centre at Madanapalle and for starting TB
Training and Demonstration Centres in
different States. He initiated the National
Sample Survey in 1956-57 and a number of
Dr. P. V. Benjamin research programmes.
Born on 21st January, 1896, Dr. P.V. Dr. Benjamin conducted in 1943 a Survey
Benjamin graduated from Madras Medical of sanatoria in India with a view to provide
College in Medicine in 1922. After taking his accommodation for TB patients in the Indian
Diploma in Tuberculosis Diseases at Cardiff Army. As a member of the TB Sub-Committee
in 1930-31 he toured extensively the Scandi- of the Health Survey and Development Com-
navian countries, Europe, Germany, the U.K. mittee he submitted the Memorandum on
and the U.S.A. to study the anti-tuberculosis tuberculosis. He was a member of the Health
work in those countries. He was Panel of the Planning Commission. He was
attached to the Union Mission TB Sana- closely associated with the Indian Council of
torium from 1922 as one of its senior doctors Medical Research. He contributed a Section
and was its Medical Superintendent for over on “Tuberculosis in the Tropics” in the book
ten years. He became Medical Commissioner “Symposium on Tuberculosis” written by Prof.
of the Association in 1941 and in 1944 was Heaf of Cardiff.
designated as its Technical Adviser. In 1948
he was appointed as TB Adviser to the Gov- Dr. Benjamin is well-known among the
ernment of India and continued functioning as international group of TB workers. He was a
Technical Adviser to the Association. He delegate to the first Empire TB Conference
retired from Government service in October, held in London in 1937 where he presented a
1962 but continued as Technical Adviser to the paper on ‘Indian People and the TB Problem’.
Association upto the 1st of July, 1964. He attended almost all International Confer-
ences in TB subsequently. He was closely
Dr. Benjamin is an ardent champion of associated with the International Union and
voluntary work and did his best to promote was a member of its Executive Body, Council
the work of the TB Association. He was and Technical and Programmes Committees
member of the Technical Committee of the for several years. He was President of the
Association from its inception in 1948 upto International Union during 1955-57 and
1964 and was President of the Nineth Tuber- presided over the XIVth International TB
culosis Workers’ Conference held in Lucknow Conference held in New Delhi in January,
in 1952. He was Editor of the Indian Journal 1957. He visited several countries in the East
of Tuberculosis from 1953 to 1964. He was in connection with the formation of the
responsible for introducing the TB Seals Sale Eastern Regional Committee of the Interna-
Campaign in India, for developing the Mehrauli tional Union and was its President from 1957
TB Hospital and for upgrading the New Delhi to 1964. He was a member of the W.H.O.
TB Clinic as a Training and Demonstration Expert Committee on Tuberculosis for a
Centre. He was responsible for the introduc- number of years and was consultant to the
Ind. J. Tub., Vol. XVI, No. 2
W.H.O. Seminar on Tuberculosis held in Dr. Benjamin is regarded as the ‘Father’
Sydney in May, 1960. He represented India at of the anti-tuberculosis movement in India and
the South East Asia Regional meeting of the as an ‘Elder Statesman’ among international
W.H.O. held in Djakarta, Indonesia, in 1955. experts. He was awarded the Kaiser-I-Hind
He advised the Nepal Government in regard Gold Medal in 1945. He was awarded in 1955
to their anti-tuberculosis work. He has pub- TADMA SHRF by the President of India
lished over 100 papers on survey, research and and the Sir Robert Philip Gold Medal by the
various aspects of tuberculosis. He was made N.A.P.T., London. In recognition of his
Honorary Life Member of the International outstanding services the Tuberculosis Associa-
Union Against Tuberculosis at its meeting in tion of India is honouring him with its Gold
Amsterdam in 1967. Medal-1969.
Ind. J. Tub., Vol. XVI, No. 2
PLANNING OF RESEARCH STUDIES*
(Some General Considerations)
S. S. NAIR
(From National Tuberculosis Institute, Bangalore)
Need for a perspective on research be necessary to find a speedy solution to this
Research is a word which is commonly
used now in many walks of life. But it is What is research ?
doubtful whether it has a commonly accepted
meaning among medical research workers or Research has been defined by Bernard Ostle
even among the smaller group of tuberculosis as an inquiry into the nature of, the reasons
and chest diseases workers. Some have their for, and the consequences of any particular set
own pet idea of research and are indifferent of circumstances—whether these circumstances
to or even cynical about other forms of re- are experimentally controlled or recorded just
search. It is important to accept a general as they occur. To be of value, research must
definition of research and to recognise that all provide reproducible results which can be
types of research are valuable and also that extended to more complicated and general situ-
their relative importance may change from ations, at least to a limited extent. Research
time to time. The progress of research can be has also to be considered as a continuous pro-
substantial and real only if research activities cess involving the following stages:
are guided with a proper perspective of such
changes in the relative importance of the vari- 1. Careful study of available data to for
ous types of research. A disproportionate em- mulate hypotheses to be tested in a new
phasis on certain out-moded forms of research study,
can prove to be very detrimental to speedy
scientific developments. This dynamic appro- 2. Designing a proper experiment or
ach necessarily implies that a small group of study to test these new hypotheses,
experienced research workers, with proper 3. Collection of data by careful experi-
perspective of the trends and future needs of mentation or observation,
research should provide the necessary guidance
and co-ordination to a band of research wor- 4. Testing the new hypotheses on the
kers with appropriate attitude towards basis of the data collected, and
research. 5. Careful study of observations which are
not in agreement with accepted hypo
Attitude towards research theses and formulation of other hypo
The importance of a proper attitude towards theses, if necessary.
research has been emphasised by many eminent
scientists. In 1949, Sir George Pickering, Any research study could be expected to
Regius Professor of Medicine in the University serve a dual purpose. Firstly, it provides data
of Oxford stated that the attitude of mind to test the hypotheses or answer the questions
“tends to be ominicient rather than admit which prompted the study. Secondly, a care-
ignorance, to encourage speculation not solidly ful study of the data could result in the formu-
backed by evidence, and to be indifferent to lation of new hypotheses to test which further
the proof or disproof of hypotheses. And, it studies are required. If the first purpose is not
is above all, to this habit of mind so inimical served it is bad planning and if the second is
to scientific inquiry that the experimental not done it is a failure to take advantage of all
method has found so small a place in clinical the information which has been collected.
studies”. It will be quite instructive for each Sometimes, research workers are tempted to
one of us to ponder and judge for ourselves reject or ignore those observations which do
whether this statement, made 20 years back, not fit in with their hypotheses. This practice
still holds good, and if so to what extent. Can is not desirable. In fact, these unexpected obser-
we at least have the satisfaction that it is no vations are likely to provide golden opportuni-
longer true for the majority of medical research ties for the research worker to get closer to the
workers in India. If not, there is room for truth. Let us remember that it is the isolated
considerable improvement in this respect and light-house and not the cluster of houses that
concerted efforts over the whole country may help the sailors to reach their destination. The
* Paper read at the 24th National Conference on need for and importance of an objective scru-
TB of Chest Diseases held on Trivandrum in tiny of unexpected observations is often not
January 1969. clearly understood.
Ind. J. Tub., Vol. XVI, No. 2
38 S. S. NAIR
It is also instructive to remember that there (e.g., normal blood pressure, blood
are two clear trends in research, which are sugar level etc.)
somewhat contradictory. While there is a ten-
dency for extreme specialization on the part of 5. studies to develop diagnostic techniques
individual research workers, most research by measurement of sensitivity, specifi-
problems are such that many disciplines and city and overall accuracy, and
fields of specialization have to collaborate for
finding the best solutions to the problem. These 6. operations research to achieve maximum
two trends imply that the individual research efficiency in the application of existing
worker has to understand and maintain a proper knowledge and skill.
balance between specialized research and
collaboration in research on multidisci- The first type has been and shall always be
plinary team basis. Thus, we should not (and of fundamental importance. Based on the
probably could not) any longer refrain from foundations so provided, types 2 to 5 evolve
identifying ourselves as an army of research the technical knowledge necessary to achieve
workers, who have to collaborate sooner the cure and control of diseases. Type 6
or later so that quicker progress can be obtain- deals with all aspects of conducting or opera-
ed in both quality and quantity of research ting a system in its natural environment and
work. Such an army of research workers recognises the fact that technical knowledge is
should have a proper understanding of the only one of the components of the system. In
planning of research studies and should be these days, when the efficient management
amenable to a self-imposed research discipline of any organisation or programme, be it big
which alone can ensure research work of or small, depends not only on the level of
uniformly good quality. technical know-how but also more and more
on many operational factors, the importance
Types of studies of operations research is being increasingly
recognised in many fields. It is only a matter
Research studies could be broadly classified of time that these concepts will find easy
in a number of ways. One type of classification acceptance among medical research workers
is into prospective and retrospective studies. also. The earlier this trend is visualised and
The latter cannot generally provide such clear acted upon, the more advanced and substan-
cut and reliable answers as the former, and tial can be our contribution to the speedier
might even be misleading if not cautiously development of medicine and public health.
used. Yet, these have an important place. For
example, double events such as congenital de- A third manner of grouping research
fects following an attack of some diseases can studies is on the basis of whether the results
be studied only when these occur by chance obtained directly contribute to a practical
from time to time and cannot be planned. course of action or mainly contribute to
Similarly, we cannot submit mankind to a knowledge only. Whatever the method of
large scale smoking experiment for 20 to 30 grouping, it is important to remember that
years to measure the relative frequencies of can- many types of research studies are possible
cer of the lung among smokers and non-smo- and that the type of study to be chosen will de-
kers. Retrospective studies provide valuable in- pend on the objectives of the study.
formation in such instances. This category of
studies can also be used to formulate hypothe- Defining the Objectives
ses for further prospective studies. A clear formulation of the objectives is the
first step in planning of research studies. This
Research studies could also be grouped is also the most crucial step because every
according to the fields or aspects they cover subsequent step in the study is dependent on
viz., the objectives. For instance, the composition
of the study groups and the accuracy required
1. basic research, for the measurements or observations are
dependent on the objectives of the study.
2. controlled trials (e.g., for testing effica- Sometimes slight variations in the objectives
cy of drugs or vaccines), may have to be accompanied by vast changes
in the design of the study and may even re-
3. epidemiological and sociological sur- quire the choice of another type of study al-
veys, together. For instance, a comparison of the.
allergy inducing capacity of two BCG vaccines
4. studies to define ‘normals’ to judge can best be studied by a simple controlled
point or level for ‘some abnormality trial with a trained team of field workers. Let
Ind. J, Tub., Vol. XVI, No. 2
PLANNING OF RESEARCH STUDIES 39
the objective be changed to camparison of of the study and the size of the study popula-
allergy inducing capacity of two BCG vaccines tion to be examined or observed. It is also
under field conditions. With the addition of important to expLaln to the statistician all the
the last three words to the objective, a number aspects which are relevant or may be even
of complicating factors such as effect of varia- doubtfully relevant to the problem and the
tions between teams of technicians, differen- onus for this must rest with the research
tial effect of storage on the two vaccines etc., worker.
crop up. The problem can then be studied
only with the help of a considerably more com- Type of Study to be chosen
plicated design and may even require a series
of studies. Once the objectives of the study are clearly
In defining the objective it is essential that formulated the choice of the best type of study
the exact sense in which each term is used is becomes fairly obvious. However, practical
known and thoroughly understood. It is also considerations may restrict this choice. For
necessary that the objectives are stated as com- instance, if making fresh observations in a
pletely as possible and do not leave room for planned manner is not practicable, only re-
different interpretations regarding the context trospective studies could be attempted even if
and scope for generalisation. If the study is to a prospective study is considered to be more
form the basis for a practical course of action, suitable. Similarly, limitations of resources
the expectations in this respect should also be may also limit the choice to the second best
clearly stated. type of study. These restrictions might also
It would be a good working principle to involve a re-formulation of the objectives.
formulate in advance, on the basis of available
knowledge, as many hypotheses which can be Design of the Study and Allocation
studied with the available resources. Such an into Study Groups
open-minded approach is the essence of good
planning and prevents a dogmatic attitude. The design of a study is the complete se-
To quote Francis Bacon “If a man will begin quence of steps considered in advance to en-
with certainties he shall end in doubts, but if sure that the appropriate data will be collected
he will be content to begin with doubts he in a systematic and well defined manner so
shall end in certainties”. During the prelimi- that an objective analysis of these data could
nary stage of formulating one’s doubts or lead to valid inferences about the problem
hypotheses it is important to ask oneself ‘can under study. Design of the study should be
this hypothesis be tested by an experiment and the primary responsibility of the statistician.
if so in what manner ?’ This will help a great
deal in selecting out the hypotheses to be tested The requirements of a good design are that:
by the proposed study and thereby lead to a
clear enunciation of the objectives of the pro- 1. The comparisons to be made should
posed study. be, as far as possible, free from syste
matic error, bias and influence of the
Scope of the Study factors which cannot be separated, eg.,
if initial and follow-up X-ray pictures
Closely related to the objectives of the are read by different readers, the
study is its scope. While the objectives define change in status cannot be ascertained
what we want to find out about a particular correctly because reader differences
population, the scope is the extent to which the also play a part and cannot be separa-
findings from this population can be generali- ted. Similarly any systematic error
sed. For instance, the findings from a repre- or bias affecting one group will also
sentative sample of patients attending an urban be merged with other differences bet-
clinic can be strictly true only for patients ween this group and the other groups
attending that clinic, but can be generalised to studied.
patients at other similar urban clinics in that
city or town and may be to such patients in other 2. The comparisons should be made suffi-
urban areas also. The extent to which such ciently precisely and it must be possi-
generalisation can be made depends upon how ble to assess the uncertainty or lack of
far the particular population studied is repre- precision in the conclusions. To en-
sentative of the population for which the sure this, it is necessary to, allocate the
Results have to be generalised. Both objectives study population into different groups
and scope of the study should be thoroughly on the basis of the statistical principles
discussed with the . statistician so that he could of random sampling and estimation
assist in selecting the type of study, the design of error.
Ind. J. Tub.,. Vol. XVl, No. 2
40 S. S. NAIR
3. the conclusions should have as wide a and the data collected will therefore be based
range of validity as possible without on differing definitions. This is particularly
decreasing their precision, true when observations are recorded in a
number of centres or over a long period of
4. the experimental arrangement should time.
be as simple as possible, and
Conduct of the Study
5. there should be a reasonable balance
between precision of the conclusions During the course of the study, adequate
and the cost of the study. If the supervision should be exercised to ensure that
conclusions are not reasonably precise the work instructions are followed strictly.
the study is almost useless, but aiming This has to be visualised at the planning stage
at an unnecessarily high degree of itself and arrangements should be made to
precision implies avoidable wastage. ensure regular flow of information to a statisti-
This is so, because, other conditions cal unit which can thoroughly scrutinise the
remaining the same, the precision of data collected and report on departures from
the comparisons between any two study work instructions. This is particularly import-
groups increases or decreases as the ant in the early stages of the work and could
number of persons studied in these prove very useful to the supervisor. This
groups increases or decreases. process of scrutiny of records and supervision
should be continued throughout the study so
In studies dealing with uncontrolled popula- that uniformly good quality of data can be
tions, it is essential that very high coverage of collected. Choosing the best type of study and
the study population should be obtained. If a proper design and absolute accuracy in
this is not possible the design should include analysis and interpretation of data will be of
collection of other relevant information regard- no use at all if the basic data are incorrect or
ing the non-respondents to find out whether unreliable in any way. It is not unusual to
they form a special group. find that data are collected in a hotch-potch
manner under the wrong impression that
At the time of finalising the design of the analysis can get something useful out of it,
study, it would be helpful if the outline of the especially if a statistician can juggle with it.
type of tables (i.e., dummy tables) for analysis Adequate Provision for Analysis and
of data and the statistical tests of significance Interpretation
for proper interpretation of the data can be
visualised. This helps in ensuring, before it is At the time of planning a study, adequate
too late, that the study population and its provision should be made for analysis and
allocation according to the design are suitable. interpretation’ of data. This is particularly
Also, wastage, by collection of unnecessary important for larger studies. This aspect is
details which will not be made use of later, can often ignored till the study is completed and
be avoided. leads to considerable delay in preparation of
the report. Even more important is the
Preparatory Work continuity of the staff from planning to report-
The next important step is the preparation ing. The team of research workers (including
of forms and cards for recording of informa- the statistician) who have planned the study
tion during the field or laboratory work and and lived with it will know the pros and cons
of detailed work instructions to each category of the material very intimately and they alone
of staff concerned, on how to carry out each can do full justice to it at the reporting stage.
step of their work and record their observations.
These cards and forms as well as the work Need for Collaboration with the Statistician
instructions should be pre-tested during a pilot
phase and finalised on the basis of this experi- The foregoing paragraphs have also indicat-
ence. If necessary, further training should be ed the need for a close collaboration between
given to the staff before starting the actual the medical research worker and the statistician.
study. It is important to ensure that the It would be instructive to consider the current
definitions and terms used are thoroughly practices in this respect. A perusal of the
understood and uniformly interpreted by each current literature shows that there is an
member of the staff concerned. Definitions increasing “prevalence” of the statistical
can often be misinterpreted during actual methods in scientific studies. But, unfortunately,
observations and recording of data. Each to some extent at least, this is based, not on an
interpretation is in effect a separate definition understanding of the underlying reasons but on
Ind. J. Tub., Vol. XVI, No. 2
PLANNING OF RESEARCH STUDIES 41
the assumption that use of tables, graphs and operative studies become all the more import-
mathematical formulae give more authenticity ant. Important findings from a study in one
and respectability. It is not uncommon to find part of the country could be generalised with
reports on some truly subjective studies which more confidence, if similar studies are conduct-
have thus been invested with a false show of ed under different realistic conditions. Such co-
objectivity. The lack of proper appreciation operative studies become all the more
of the importance of statistical thinking and important in operations research because the
methods at various stages of a research study practical conditions under which knowledge
has similarly led to varying levels of collabora- and skill are actually applied for the benefit of
tion with the statisticians. Sometimes such the community or nation introduce a number
collaboration is considered as a formality to of key variables which have great influence on
ensure financial resources or to bestow a stamp the outcome. Some factors considered to be
of acceptability. There are also instances of technically important pale into insignificance
collaboration which are on a par with the in the light of some of these key variables.
following answer to the question, “Do you
believe in ghosts ?” “No, but I am afraid of Need for a Rigid Research Discipline
them”. The belief that statisticians deliberately
make simple questions difficult is also not The success of such co-operative studies
uncommon and could have retarded the growth would depend upon the participating units
of healthy collaboration between the doctors following a rigid research discipline. Even
and statisticians. Successful collaboration studies in one area only, will become more valu-
demands that the statistician should learn all able if the quality of the research work becomes
he can of the problem in question and the uniformly good and reliable. What is often
medical man should learn all he can about the not recognised is that a self-imposed research
statistical approach. Without substantial discipline is essential among research workers.
knowledge on both sides it might turn out to The more rigid it is, more valid will be the
be, blind leading the blind. Those research conclusions and more comparable will be the
workers who are reluctant to learn or apply the findings of different studies. Whereas problems
statistical approach may not have realised that vary from area to area and population to
a new language is a riddle before it is conquered population, the research discipline need not
but a power in the hand afterwards. In any vary, as the basic concepts are fundamental
case, it is important to remember that the time and uniformly applicable. The reasons for the
for consulting a statistician is before planning lack of self-imposed research discipline can,
the study, during the planning, during interim therefore, be only artificial or man-made.
review to ensure good quality of data and
during the analysis and interpretation—in fact, Conclusion
he should be involved in all stages of the study Some of the statistical principles which have
whenever possible. to be kept in mind while planning research
studies and the various steps that have to be
Need for Co-operative Studies gone through have been indicated earlier. It
would be a very revealing experience to go
The field of medical observation is compli- through the research studies already under-
cated because of some inherent variations. No taken or in progress and see how many of these
one doctor can treat sufficient number of cases satisfy these principles or have gone through
in a short span of time and a large number of the various steps listed. It is not unusual to
doctors may each treat a few cases. The cases find reports of research studies in which the
themselves are far from being a homogeneous questions or objectives are either not clearly
group. The research worker in the field of formulated or are not stated at all. Sometimes,
public health faces the further difficulty that he the type of studies chosen, the design, the study
cannot set up and control his own experiment population and the analysis are not suited to
in his own laboratory or clinic. In such provide valid answers to the questions formu-
circumstances, both for clinical and public lated or the hypotheses to be tested. Lack of
health research co-ordinated team work in the adequate arrangements for continuous super-
same area or spread over a number of areas vision and scrutiny of methods of observation
is often necessary. In the absence of such co- and recording is much more common than
operative studies the literature becomes full of normally visualised. There are also instances
cLalms, assertions and counter assertions each which remind one of the saying that “statistics
of them being correct in its own limited are used the way a drunkard uses a lamp post;
manner, like the four blind men describing the for support rather than for light”. The only
elephant. In a country like India, where practicable remedy for these situations is a self-
conditions can vary to a large extent such co- imposed research discipline.
Ind. J. Tub,, Vol. XVI, No. 2
PLANNING, CONDUCT AND EVALUATION OF CONTROLLED CLINICAL
(From Tuberculosis Chemotherapy Centre, Madras)
The controlled clinical trial is now a well- very clearly the type of patients to be admitted.
accepted method of measuring the relative effi- To give an example, Slide 2 sets out the impor-
cacies of different therapeutic regimens for tant criteria employed at the Tuberculosis
many diseases. Although its usefulness is Chemotherapy Centre, Madras.
widely appreciated, there is an insufficient
awareness of the rationale and the methodology 2. Choice of Patients for study at the T,C.C.,
of the controlled clinical trial—that is, the rea- Madras
sons underlying it and the procedures involved
in the execution. By taking examples from the
field of pulmonary tuberculosis, the isssues in- 1. Aged 12 years or more
volved can be clearly set out. 2. No previous chemotherapy
Specification of regimen and priority in aims 3. Bacteriologically confirmed pul. Tb.
To evaluate the efficacy, toxicity and accep- 4. Drug-sensitive organisms
tability of an anti-tuberculosis regimen—for 5. Bonafide residents
instance, isoniazid plus thioacetazone, it is
necessary to start with very clear ideas of the
dosage, the rhythm of administration and the It is important also to specify centra-indica-
exact duration of the regimen. Next comes tions for admission to the study—for example,
specification of the order of priority in aims as patients with leprosy or diabetes, since their
there can be a clash of interests. For instance, management would be rather complicated.
I. Specification of Regimen and Priority in Aims
The next requirement is a control group of
Need for control
(a) dosage, rhythm and duration
(b) priority in aims Slide 3 gives some interesting examples
of entirely inaccurate or highly misleading con-
(1) Efficacy clusions that one might draw in the absence of
(2) Toxicity a control group of patients.
3. Need for Control
if the main aim is to determine the efficacy of
the drugs, it will obviously be necessary to
employ procedures for detecting irregularities Rx for 3 years I%of77
in drug-collection and drug-intake, and correct- Rx for 2 years 0% of 74
ing them. Such action would, however, mean
that pressure is applied on patients when they
show evidence of non-acceptability, thereby Strep. + 1NAH 35% of 78
making any assessment of the acceptability of Giddiness
the regimen rather artificial. This is a good PAS + 1NAH 11% of 70
illustration of the basic maxim that any study
can have only one main aim. The first example refers to relapse rates in
Choice of patients for study at the T.C.C., patients with bacteriologically quiescent tuber-
Madras culosis. In patients who received chemotherapy for
3 years, the total relapse rate in the third, fourth
For any generalisation to be possible from and fifth years was only 1%. This low proportion
the results of a study, it is necessary to define could have led to the recommendation that 3 years
of chemotherapy is absolutely necessary to keep
* This paper was presented as a lecture at the 24th the relapse rate low. (Indeed, similar
National Conference on TB and Chest Diseases, held recommendations have been made in the
in Trivandrum in Jan. 1969 literature, in the absence of controls). Such
Ind. J. Tub, Vol. XVI, No. 2
PLANNING , CONDUCT AND EVALUATION OF CONTROLLED CLINICAL TRIALS 43
a recommendation is, however, totally unwarn- standards for smears, cultures, sensitivity tests
ted since, in the patients who received only 2 and urine tests. Obviously, the only way out of
years of chemotheray, the relapse rate was these dangers is to have a control group that
0%. is concurrent.
The next example is a less extreme one, and
pertains to toxicity. In a group of patients Number of patients to be admitted
treated with a twice-weekly regimen of strep-
tomycin plus isoniazid, 35% compLalned of Next, let us consider the question which is
giddiness on at least one occasion during the most frequently posed to the statistician, namely,
year of chemotherapy. However, 11 % of the “How many patients must I admit to the study
control group (who received a standard regi- to obtain a statistically valid result ?” Unfor-
men of PAS plus isoniazid) also compLalned of tunately, the short answer to this question is
giddiness. Thus, in the absence of the control that there is no such magic number. However,
group, we would have acquired an exaggerated if the clinician can indicate to the statistician
picture of streptomycin toxicity. the approximate efficacy of the control regimen
and, furthermore, state what difference from
Examples like this are plentiful. For ins- the control regimen he would regard as having
tance, in the treatment of tuberculosis, conclu- practical importance, the statistician can then
sions about the value of gold therapy, value of tell them approximately how many patients
hospitalization and role of diet have been should be admitted.
drawn and, in the case of the latter two, are
still being drawn without having a control To take an example, the clinician might be
group of patients. interested in the new regimen only if it is 20%
These examples will have convinced you of more effective than the control regimen, which
the necessity for having a control group of from previous experience is known to have an
patients. In the present context, the control efficacy of 75%. In this case—that is, an effi-
might be a regimen that is already in use at cacy of 75% for the control and 95% for the
your clinic, for instance, a standard regimen of new regimen, approximately 70 patients will
isoniazid plus PAS. have to be admitted to the study (that is, 35 in
each series) to demonstrate statistical signifi-
Need for concurrency cance. If, however, the clinician wishes to
4. Need for Concurrency 5. Number of Patients to be Admitted
Factors that could vary Control New No. of patients
regimen regimen to be admitted
1. Disease condition of patients
75% 95% 70
2. Co-operation of patients
3. Clinic supervision 75% 90% 130
4. Laboratory standards 75% 85% 290
75% 80% 1150
Next, it is essential that the control
should be a concurrent one. Comparisons with
a non-concurrent control—that is, retrospective detect a smaller degree of superiority, say 15%
comparisons—are usually dangerous, as there (that is, an efficacy of 90% for the new regi-
are many factors that could vary from one point men), the number required will be 130. The
in time to another. For instance, the disease corresponding number for a 10% superiority
condition of the patients admitted to will be 290, and fora 5% superiority 1150.
treatment might be different in different years, Thus, the smaller the difference to be detected,
on account of changes in diagnostic measures the larger will be the number of patients
or influence of mass propaganda campaigns. r e q u i r e d . I t mu s t b e n o t e d t h a t t h e
The co-operation displayed by the patients number required for statistical significance will
might also vary from one year to another, depend not only on the size of the difference
possibly due to socio-economic causes. Thirdly, to be detected, but also on the absolute levels
the intensity of examination and the overall of efficacy of the two regimens.
quality of the clinic supervision might be
different, especially if there have been changes
in the personnel. A similar problem can arise It is worth stressing at this stage that statis-
with the laboratory tical significance need not be the sole criterion
Ind. J. Tub., Vol. XVI, No. 2
44 S. RADHAKRISHNA
for determining the number of patients to be rapy in myocardial infarction (quoted by
admitted. Very often, we are just as interested Truelove), the system of alternation resulted
in obtaining as precise an estimate as possible in 580 treated patients and only 442 control
of the efficacy of the new regimen. Obviously, patients, a difference that could have occurred
the larger the number of patients admitted, the by chance in only 1 of 5,000 occasions.
more precise will be the estimate. For instance, The best protection against all accusations
if the efficacy was found to be 80% in a sample of bias is random allocation from sealed
of 100 patients, it may be stated, with 95% envelopes. This procedure may be regarded as
confidence, that the true efficacy lies within the equivalent of tossing a coin. In practice, it
80±8%. ‘If, however, the efficacy of 80% had consists of preparing a treatment regimen list
been observed in a larger sample of patients, for successive patients based on random num-
say 400, the limits will naturally be narrower, bers that are available in statistical tables and
namely, 80±4%. incorporating it into sealed envelopes. Each
Summing up, the decision regarding the sealed envelope must have written on its exte-
number to be admitted must be based on objec- rior the name of the study, and a sequential
tive considerations like statistical significance serial number. Inside each envelope, there
and high precision. Practical considerations should be a slip of paper giving the name of
like availability of patients, drugs and facilities the study, the sequential serial number and the
are no doubt important but should always be regimen for the patient. When a patient is
regarded as secondary. found suitable for admission to the study, his
treatment regimen is to be determined by
Mode of deciding the regimen for individual tearing open the next in the series of sealed
Next comes the mode of deciding the treat- Purpose of random allocation
ment regimen for individual patients. In a
controlled clinical trial, the mode of deciding The purpose of random allocation is to
the regimen for any individual patient must avoid personal preferences in the choice of
not only be free of bias, but also appear to be treatment for individual patients. It has to be
free of bias. One can readily see the danger in emphasised that these personal preferences can
entrusting the choice of the regimen for indi- be conscious or, more often, sub-conscious.
vidual patients to the clinician. To take a simple
7. Purpose of random allocation
6. Mode of deciding the Regimen for individual
1. To avoid personal preferences,
1. Clinician’s choice conscious or sub-conscious
2. Alternation 2 To construct two groups similar in all aspects
3. Random allocation from sealed envelopes (a) known and measurable (stratification)
(b) known but immeasurable
example, if patients were to be treated at home (c) unknown
or in sanatorium at the clinician’s discretion
(which might be, in some instances, influenced
by the patient’s wishes), it is almost certain that Failure to recognise that there is such a thing
the iller patients would tend to be admitted to as sub-conscious bias has often led investigators
sanatorium while the less ill patients would be to regard random allocation as a slur on their
treated at home. personal honesty.
Another highly undesirable procedure is the The great advantage of random allocation
method of alternation, whereby the first patient is that it is highly likely to result in the cons-
is prescribed regimen A, the second regimen B, truction of 2 groups which are similar in all
the third regimen A, the fourth regimen B, and aspects—known and measurable, known but im-
so on. A variant of this is to admit all pati- measurable or not measured, as well as the un-
ents on odd days to the regimen A and those known. In the case of known and measurable
on even days to the regimen B. Such proce- characteristics that have prognostic importance,
dures are, however, capable of bias because the a further precaution would be to stratify the
order in which patients are admitted to a study patients into 2 or more groups—e.g. non-cavita
can be manipulated without much difficulty. ted and cavitated—and undertake the allocation
For instance, in a study of anticoagulant the- from separate series of sealed envelopes, one for
Ind. J. Tub., Vol. XVI, No. 2
PLANNING, CONDUCT AND EVALUATION OF CONTROLLED CLINICAL TRIALS 45
each group. In the present example, this detail—that is to say, no deviations can be
procedure will ensure that the two series have made to suit the needs of individual patients or
identical proportions of cavitated patients. individual clinicians.
To facilitate strict adherance to the protocol,
Similarity in subsequent management it is useful to have the important aspects (e.g.
criteria for eligibility to study, intensity of the
Next, it is important to ensure similarity in x-ray and sputum examinations, weight-dosage
the subsequent management of the patients in schedules) abstracted on to separate sheets of
the 2 series. For this, it is necessary to set out paper that are readily available to the clinicians
in advance (1) the intensity of examination and nurses ; also, diaries for reminding clinic
during treatment—clinical, x-ray, sputum etc., staff of ensuing examinations should be kept.
(2) the nature and frequency of checks on drug- These are what can be termed as reminder
8. Similarity in subsequent management systems. Despite these, deficiencies can occur ;
it is therefore necessary to have systems for
1. Intensity of exam.—clinical, x-ray, sputum etc.
detecting deficiencies and rectifying them before
it is too late.
2. Checks on drug-regularity Well-designed forms and analysis cards
make analyses easy ; therefore much time
3. Defaulter action must be spent on them at the design stage.
4. Observance of toxic symptoms Also, information collected should be abstrac-
ted periodically on to analysis cards. This
5. Criteria for withdrawal from study will not only facilitate interim analysis, but also
highlight deficiencies in the forms, cards and
recording systems, which can then be rectified.
regularity, (3) procedures for dealing with As bias can creep in to laboratory investi-
defaulters, and (4) procedures for the recording gations, it is important to devise systems in
of symptoms of toxicity. Finally, and most which there is not even scope for bias. For
important, the circumstances under which a instance, when smears are examined, or cultu-
patient may be withdrawn from the study must res or sensitivity tests read, or urine tests under-
be stated very clearly. For instance, the criteria taken, it should be arranged that the laboratory
could be serious radiographic or clinical dete- technicians are unaware of the source of indi-
rioration in the presence of a positive sputum of vidual specimens.
major drug-toxicity. It must be emphasised that Finally, it is essential to have quality con-
all these procedures must be implemented alike trol for laboratory tests and for drugs that are
for all patients, regardless of the treatment in use in the study. At the Tuberculosis Chemo-
regimen. therapy Centre, we keep track of the standards
Conduct of the study in the laboratory investigations by slipping
in controls without prior warning and by perio-
All that has been said so far relates to the dic reviews of the incidence of contamination
planning of a controlled clinical trial. When and smear-positive culture-negative results. As
the plan is fully evolved, a protocol should be regards drugs, assays are undertaken routinely
written up which contains all these points, and on arrival, and if necessary at periodic intervals
made available to all participating physicians. thereafter.
The protocol should be treated as a sacrosanct
document, and scrupulously observed in every Evaluation of results
9. Conduct of the study Even in the case of well-planned and well-
10. Evaluation of results
1. Strict adherance to protocol
Reminder systems 1. Be wary in excluding patients from analyses
Deficiency-detecting systems 2. Check for similarity between series in
(a) initial condition
2. Design of forms and analysis cards
(b) intensity of examination during treatment
3. Periodic abstraction of information
3. Objective methods to ensure bias-free comparisons
4. Avoidance of bias in lab. investigations (a) Independent assessor for x-ray reading
5. Quality control—lab. tests, drugs (b) Clear definitions of fav. and unfav. response
Ind. J. Tub., Vol. XVI, No. 2
46 S. RADHAKRISHNA
conducted studies, great care has to be taken in ther, the x-rays should be fed to the assessor
the evaluation of the results. One common in strict sequence of the patient serial number,
error is the exclusion of patients from final which is by design a random sequence. Defini-
analyses. Sometimes, the reasons are obviously tions of favourable and unfavourable response
unrelated to the treatment regimen; in such must be clear-cut, and applied alike to all the
cases, it is sufficient to establish that the exclu- patients regardless of the treatment regimen. In
sions have occurred to a similar extent in both other words, classifying patients as having a
series. However, we have had examples at favourable or unfavourable response on an
previous conferences where deaths from tuber- individual basis without laying down strict
culosis were conveniently excluded and cheer- definitions is a highly objectionable procedure.
fully optimistic conclusions drawn from the
findings in the survivors. Such procedures must No evaluation can be complete without
be deplored strongly. The rule should be to tests of statistical significance. However, the
describe the progress of all patients admitted results of these tests must not be regarded as
to the study who belong to the population giving proof of existence or proof of non-exis-
defined earlier (Slide 2). tence of a difference. Thus, when we say that
a difference is statistically significant, all that
Although random allocation can be expec- we mean is that the likelihood of it being a
ted to yield 2 series which are very similar in fluke observation less than 5%.
their initial condition, nevertheless, analyses
should be undertaken to check that the 2 I would like to stress that planning, conduct
series were in fact similar on admission. Also, and evaluation are not three water-tight com-
analyses should be undertaken to check partments that can be dealt with independently
that the actual intensity of examination during by different people or different committees.
treatment was the same. Atleast one individual, preferably the chief
investigator, must be deeply involved in all
Finally, for assessing x-ray progress, it is three stages, and all the other participants must
important to obtain the services of an indepen- understand and appreciate the rudiments of
dent assessor who is not connected with the controlled experimentation, if the outcome of
day-to-day management of the patients. Fur- such efforts is to be valuable.
Ind. J. Tub., Vol. XVI, No. 2
PLANNING AND CONDUCT OF EPIDEMIOLOGICAL SURVEYS*
G. S. ACHARYULU
(From Madanapalle Tuberculosis Research Unit, Madanapalle)
Pulmonary Tuberculosis is the most com- characteristics. In the latter situation, a sample
mon form among the different forms of tuber- reasonably representative of the population is
culosis. The main indices useful in the selected according to practical convenience
epidemiological study of this disease are the for undertaking the survey. In the former
prevalence and incidence rates of (i) persons situation, the population will have to be
infected with tubercle bacilli, (ii) persons divided into a number of strata and a repre-
excreting tubercle bacilli and (iii) suspect cases sentative sample will have to be selected from
of tuberculosis (diagnosis not confirmed each stratum. An example of such a situation
bacteriologically). With the occurrence of is provided by the presence or absence or non-
primary drug-resistant cases, knowledge of the specific sensitivity in the population in a study
prevalence and incidence of such cases has designed to investigate the protective effect
attained considerable importance. of BCG Vaccination, as there is reason to
believe that non-specific sensitivity offers some
There are two types of epidemiological protection against tuberculosis.
surveys, prevalence surveys and longitudinal
surveys. In prevalence surveys, the observa- Special problems which are to be consi-
tions refer to a specific point in time. In dered in planning experimental surveys are :
longitudinal surveys, the observations are repea-
ted at different points in time. Both prevalence (i) incorporation of suitable ‘controls’,
and incidence rates can be estimated from (ii) method of randomization and
these surveys. They are useful for making
future projections of the tuberculosis problem (iii) determination of the size of the experi-
and in evaluating the effects of different mental population.
tuberculosis control programme.
In situations, when there is doubt regarding
Longitudinal surveys can be classified as the effects of T.B. control measures or when
experimental surveys and non-experimental the interest is in evaluating their effects,
surveys. The objective in experimental survey ‘controls’ must always be introduced in the
is to assess the value of or to compare the experiment. If, on the other hand, there is no
effects of different T.B. control programmes. doubt regarding the effects of control mea-
In non-experimental surveys, no such assess- sures and the interest is only in comparing the
ments or comparisons are contemplated. effects of different control measures, there is
Considerations in Planning the Surveys no necessity for ‘controls’. If valid compari-
sons are to be made, the experimental units
The first step in planning an epidemiologi- (an experimental unit consists of a single
cal survey is to define the objectives of the individual or any well defined group of indivi-
survey. This involves the specification of the duals) should be randomly allocated to the
purpose of the survey and the epidemiological different programmes. There are many methods
indices which are to be estimated, and how the of randomization and they make use of the
survey results are going to be made use of. knowledge, if available, of the inherent vari-
On the basis of these indices it will be deter- ability of the experimental material which is
mined whether the survey should be a preva- relevant to the investigation. The purpose of
lence or a longitudinal type. these methods is to increase the efficiency of
the experiments by reducing the experimental
In experimental survey, the selection of the error. The factors to be considered in the
individuals into the study depends upon the choice of a proper method of randomization
type of population to which the experimental are :
results are to be later applied. There are two
situations : the first is, the population is sup- (i) the types of control programmes to be
posed or known to possess certain characteris- compared,
tics which will interact, favourably or unfavoura (ii) available knowledge of the inherent
bly, with the T.B. control measures, which are to variability of the experimental material
be compared in the experiments ; the second is and
when there is no knowledge about any such
(iii) practical considerations in the conduct
* Paper read at the 24th National Conference on TB of the survey such as actual process of
and Chest Diseases held in Trivandrum in randomization etc.
Ind. J. Tub., Vol. XVI, No. 2
48 G. S. ACHARYTJLU
Size of the experimental population will be (i) reduction in the cost of the survey,
determined by making use of the estimates of
of experimental errors obtained from previous (ii) the speed with which the estimates
surveys. can be obtained and
(iii) a greater accuracy in the results.
There are two methods of undertaking
prevalence surveys or non-experimental Greater accuracy will be obtained as it will be
surveys ; complete survey method and sample possible to control the various sources of errors
survey method. The objective in both the in collecting the data by employing technically
methods will be to obtain knowledge of the trained and experienced personnel, when the
distributions, aggregate or mean values of one survey is conducted on a limited scale.
or more characteristics of a well defined popu-
lation. In complete survey method, observations The undertaking of sample survey requires:
are made for every individual in the total
population, whereas, in sample survey method, (i) the construction of a sampling frame,
observations are confined to a certain number
of individuals in the population. On the basis of (ii) the choice of a sampling design and
the selection of the individuals, samples can be (iii) the fixation of the sample size.
divided into random and judgement samples.
In random samples the individuals are selected The sampling frame consists of a list of sam-
according to the laws of probability while, in pling units in the total population without
judgement samples, selections are not done omission and/or duplication and this is essential
according to these laws. The National Tuber- for drawing random samples. The sampling
culosis Sample Survey, conducted during unit may consist of an individual or all members
1955-’58, under the auspices of the Indian in the household or a group of households in
Council of Medical Research, is an example of a contiguous area or any other well defined
random sample survey. The European group. The sampling design stipulates the
Tuberculin Survey undertaken during 1965-’66 method of drawing the sampling units into
under the auspices of the International Union random sample. For the same sample size,
Against Tuberculosis and the First Drug- certain sampling designs will be more efficient
resistance Survey (1964-’65) conducted under than others, efficient in the sense of containing
the auspices of the Indian Council of Medical comparatively less sampling error. Although
Research are two examples wherein judgement it is possible to minimize the sampling errors by
samples have been made use of. choosing efficient sampling designs and suitably
increasing the sample size, in practice, the deci-
The random sampling method makes it sion regarding the sampling design will have to
possible not only to estimate the population be taken on the basis of the sampling frame
characteristics but also to estimate the errors that is available and mainly in view of the
involved in those estimates. This method will operational convenience of the field work. The
ensure that the samples will be representative size of the sample will be fixed by the consi-
of the population. On the contrary, there is deration of the cost of the survey and the
no guarantee that the judgement samples will margin of error allowed in the estimates.
be representative of the population. Further,
the errors involved in the estimates of the Types of Data collected
population characteristics cannot be estimated
from judgement samples. Another disadvant- Specific data to be collected in any particular
age is, that it will not be possible to make survey will be decided by the specification of
valid comparisons of the estimates obtained the layout of the final statistical tables required
from one judgement sample with those from the survey. The types of data that are
obtained from another. In view of these con- generally collected in epidemiological surveys
siderations, random sample surveys are always relate to (i) census, (ii) tuberculin tests, (iii) X-
to be preferred. Judgement samples should ray examinations, (iv) bacteriological exa-
be made use of only if it is not possible, owing minations, and (v) anti-tuberculosis measures.
to practical difficulties, to undertake random A discussion of the considerations in collecting
sample surveys. some of these data may be relevant here.
With regard to the choice between random Tuberculin Tests
sample surveys (hereafter referred to as sample
survey) and complete surveys, sample surveys Although the importance of this test in
are always to be preferred as they have several measuring the prevalence and incidence of
advantages over complete surveys. These are : infection is well known, its value is restricted
Ind. J. Tub., Vol. XVI, No. 2
PLANNING AND CONDUCT OF EPIDEMIOLOGICAL SURVEYS 49
by the use of BCG Vaccination. Hence, this as census taking, tuberculin testing
factor, that is, the extent to which the popula- etc.
tion concerned was covered by BCG Vaccina-
tion is to be considered before deciding to use The instructions to the field staff should be
this test for measuring the epidemiological unambiguous and should contain typical
indices. illustrations. If necessary, a pilot survey may
be undertaken. Such a survey will be useful
X-ray examinations (i) in the development of the field procedures,
(ii) for testing the suitability of the various
In countries like India, children in the cards and forms designed and (iii) in giving
younger age groups constitute a considerable training to the field staff.
proportion of the total population. In view of
the low prevalence and incidence of the disease An important aspect, particularly in the
among children, it is sometimes desirable to conduct of experimental surveys, must be
exclude them from these examinations. Such a mentioned here. This is to ensure that groups
procedure will reduce the work-load consider- of individuals allocated to different control
ably and at the same time the resulting loss in measures to be compared are treated alike
precision of the estimates will be negligible. thoughout the survey with regard to the admis-
Regarding the interpretation of X-ray films, a sion of the individuals into the groups, inten-
suitable code to meet the specific requirements sity of the diagnostic examinations and other
of the survey will have to be devised, as no factors except the actual control measures.
internationally accepted code* has so far been How any such factor can introduce bias into
evolved for this purpose. Two independent the study may be illustrated by an example. In
readers are considered essential. a BCG Trial, if places injections are not
given to ‘controls’, it will not be possible to
Considerations in the conduct of the Survey identify the group among the controls corres-
ponding to the group of persons who refused
For the success of the survey, it is essential BCG Vaccination. Another important aspect
to have well trained and experienced personnel. is that the identity of the control measures
When such personnel are not available, recruit- given to an individual should not be revealed
ment and training programmes will have to be either to him or to those who will have to
undertaken well in advance. The various cards assess the effects of control measures.
and forms to be used for collecting the data Sources of Errors and Measures of Control:
should be designed. A detailed protocol for
the conduct of the survey should be prepared. In order to get valid and reliable estimates
This protocol should: it is essential to minimize the errors which arise
(i) Specify the purposes and objectives from several sources. These are broadly divided
of the survey; into sampling and non-sampling errors. Sampl-
ing errors occur only in sample surveys, whereas
non-sampling errors occur both in sample and
(ii) define the populations to be complete surveys. While the sampling error, as
surveyed; mentioned earlier, can be kept within predeter-
mined limits by suitable measures, it will not
(iii) describe the various methods of generally be possible to identify all the sources
examinations to be conducted along of non-sampling errors and to estimate the
with the criteria of classifications to extent to which they affect the final estimates.
Non-sampling errors arise on account of
(iv) define the concepts to be used such several factors and some of them are discussed
as case of tuberculosis etc.; below :
(v) contain detailed instructions to the 1. Failure to examine the complete study
staff working in different sections such population : This is called the error due to non-
response. One method suggested to obtain
Efforts are being made by a specially appointed unbiased estimates of the population characte-
international committee (The Ad Hoc Commi- ristics is to examine a random sample of
ttee for the study of Classifications and Termi- persons who failed to report for the examina-
nology in Tuberculosis under the auspices of
IUAT in cooperation with WHO), to evolve a tions earlier. This procedure will not be
code for interpretation which will be acceptable practicable in many situations. The other way
internationally and which will minimize the of reducing this error is to minimize the percen-
errors of interpretation.
Ind. J. Tub., Vol. XVI, No. 2
50 G. S. ACHARYULU
tage of non-response by instensive efforts and to used for this --purpose, provided that other
note down the reasons for non-response, so that conditions are the same, comparison of esti-
an analysis can be made later to give an idea mates from one group with another will be
as to how the estimates are going to be affected. valid. But, when the interest is in estimating
the absolute value of the magnitude of infection
2. Biases : There are two types of biases, with tubercle bacilli, efforts should be made to
systematic and unsystematic. The occasional evolve an objective criterion for this purpose.
failure to detect every bacillary case with single
spot sputum specimen is an example of syste- 4. Errors in data processing and tabulation:
matic bias. Such a procedure will lead to These errors can be reduced by giving clear
consistent under-estimation of the number of instructions, by employing well trained staff
bacillary cases. The extent of under-estimation and by exercising constant supervision over
can be ascertained by conducting special studies them.
or from the results of such studies made
elsewhere. If the resources permit, this error To sum up, the first step in planning a
can be reduced by better methods of sputum survey is to provide a clear and unambiguous
collection. statement of its purposes and objectives.
Planning requires consideration of all important
The effect of unsystematic bias is sometimes aspects of the survey, many of which are inter-
over-estimation and some times under-estima- dependent. The application of statistical
tion. The unconscious preference for round methods is essential to get valid and reliable
numbers, generally observed in reading the estimates and for making valid comparisons.
tuberculin reactions and in reporting the ages, Recognition of the types of errors which might
may be given as examples of unsystematic occur in survey will not only lead to better
biases. The effects of these errors can be planning and execution of the survey, but also
reduced to some extent, by suitably grouping the to better understanding of the reliability of the
observations in preparing the frequency distri- estimates obtained from the survey.
bution when the data are tabulated.
3. Errors in classification : Generally, 1. WHO Expert Committee on Tuberculosis, 8th
tuberculin positive reactors who are supposed Report, 1964 WHO Technical Report Series,
to have been infected with tubercle bacilli are No. 290.
denned to be those whose reactions exceed a
certain size, for example 6 mm. or 8 mm. The 2. Protocol of the Coordinated European Tuber
dividing point between negative and positive culin Survey, Bull. WAT, 1965, No. 1, Vol. 36,
reactors varies from one study to another and
maybe anywhere between 6 to 15 mm. This 3. Report of the Committee on Epidemiology and
type of definition is known to involve two Statistics, Bull. WAT, 1966, No. 1, Vol. 38, 71-74.
types of errors, viz., inclusion of persons who
have been actually infected with tubercle 4. Gangadharam, P.R.J., Proceedings of the 21st
bacilli and exclusion of persons who Tuberculosis & Chest Diseases Workers’
have been actually infected with tubercle Conference, 1966, 104-112.
bacilli. The magnitude of these two types of
errors depends upon the magnitude of the 5. Report of the Ad Hoc Committee for the study
prevalence of non-specific sensitivity in the of Classification and Terminology in Tuberculo
community. As long as the same definition is sis, Bull. WAT, 1965, No. 1 Vol. 36, 55-72.
Ind. J. Tub., Vol. XVI, No. 2
ANALYSIS AND PRESENTATION OF DATA*
G. P. M ATHUR
(From New Delhi Tuberculosis Centre, New Delhi)
Assuming that a clinical trial, laboratory valid if experiments, surveys etc. have been
experiment, or epidemiological survey has been properly designed and all the rules observed.
concluded and a mass of reliable data collected, The most important are the ones concerned
there arises the question of analysis, interpre- with random allocation and the elimination of
tation and presentation. The vast quantity of bias ; and with advance planning they are not
information has to be numerically summarised difficult to comply with.
and logically tenable conclusions drawn there- Statistical data usually are affected by a
from. The statistical techniques for dealing multiplicity of causes and this is more so in the
with data are determined largely by the plan of case of biological data. For drawing valid
the study and the nature of the information conclusions from data influenced to such a
collected. Analysis, in the main, is a job for marked degree by variation and chance, there
the statistician and, as a rule, doctors and exist powerful statistical tools. These are based
medical research workers need not aim at mas- on laws of probability and that is why conclu-
tering the statistical methods applicable to sions drawn thereby, though not applicable to
different situations However, an understanding individuals, are true over large numbers of
of the genera] principles does help in the sys- individuals. A special advantage of the statisti-
tematic compilation of the right type of data, cal technique is that it furnishes estimates of
in intelligent discussion with those who are the precision of the results obtained.
mainly concerned with the statistical evaluation It would be obviously impossible within the
and in interpreting the final results. Indeed, it short space of this paper to describe or even
is only a continuation of the process, starting outline briefly the various forms of statistical
at the planning stage, whereby the medical treatment of data pertinent to different situa-
research workers and statisticians learn to tions. As already stated, these are to a large ex-
understand each other’s language and arrive at tent determined by the nature of the studies. I
a common basis for discussion. would therefore content myself with a few gene-
It would be useful to start at the point ral remarks.
where data are being collected and compiled. Interpretation of biostatistical data is not
Except for large scale surveys, it is not really merely a question of applying certain mathe-
essential in most cases to resort to mechanical matical formulae and obtaining a result.
sorting and tabulation. Indeed, for studies Even more important than these are an exercise
involving a few hundred cases it is sometimes of one’s logical and critical faculties. Medical
more advantageous to have either individual journals are replete with instances where the
cards or even simple lists. The latter, prefer- wildest conclusions have been drawn by apply-
ably prepared in code, and incorporating all ing statistical techniques contrary to the dicta-
information, are in fact to be particularly reco- tes of commonsense and elementary logic. Even
mmended for small scale studies, since they such apparently simple things as percentages,
permit an intimate acquaintance with the data, prevalence and attack rates can be dangerous
suggesting possible relationships between vari- in the hands of persons not accustomed to look
ous characteristics and, in general, indicating at figures in a critical way. Special care needs
the lines along which data should be analysed. to be exercised when conclusions are being
The form in which these entries are to be drawn on the basis of a comparison between
made itself needs to be carefully designed, two groups. Unless it is ensured that the
special attention being paid to such matters of groups are similar in all respects, except the
detail as units to be used, grouping of data one under study, conclusions will not be valid.
and and above all to comprehensiveness. In many cases, the methods of random alloca-
Simplicity, precision and legibility and other tion would have taken care of such differences.
criteria which should be kept in mind. In others one may have to resort to ‘standardi-
The importance of associating the statisti- sation’ to nullify the effect of inequalities in
cian with any research project from the very important respects.
start has already been stressed. Data collected Tests of significance are so widely used by
without any plan or with a defective plan may statisticians and so often misunderstood by
not lead to any worthwhile conclusions. Tests others that it is probably useful to say a word
of significance and statistical analysis are only about these. As an example, let us consider a
*Paper read at the 24th National Conference on simple clinical trial in which a group of 100
TB & Chesl Diseases held in Trivandrum in Jan. 1969. patients has been divided at random into two
Ind. J. Tub., Vol. XVI, No. 2
52 G. P. MATHUR
sub-groups, those in the first group getting the nineteen forties were to be compared with
drug regimen A and those in the other, drug those of a similar sample from recent years
regimen B. Assume further that at the end of a by a panel of readers. Some x-rays in the
certain period, sputum conversion rates of former sample could not be traced, which
70% and 80% have been observed in the two fact tentatively was ascribed to faulty storage.
respective groups. How are we to know whe- Since no bias was suspected, it was decided to
ther the observed difference represents a real go ahead with the comparison of the available
superiority of regimen B over regimen A or is x-rays. This revealed a rather surprising
merely due to chance ? This is done by using a finding: that contrary to the general impression
test of significance. To start with, we set up there were not many cases with far advanced
what is known as the ‘null hypothesis’. In our disease in the older series. Further scrutiny
particular example, for instance, we begin by brought out the fact that in the 1940s, it was
assuming that the two drug regimens do not in not considered worthwhile in the New Delhi
fact differ in respect of sputum conversion. We TB Centre to take x-rays of cases with very far
then calculate the mathematical probability of advanced disease. Because of the wartime
getting a difference at least as big as the one scarcity of x-ray films, and the comparatively
observed if the ‘null hypothesis’ was correct. If poor prognosis of such patients, they were
this probability is sufficiently low, say 5% or usually diagnosed by fluoroscopy and sputum
less, we reject the null hypothesis and conclude examination, no permanent x-ray record being
that the observed difference is in fact real. It considered necessary. Obviously, further
must however be remembered that, based on analysis of this data had to get round this
the laws of probability as it is, the test of signi- difficulty. Pitfalls like this are not at all rare
ficance cannot be said to have ‘proved’ the in analysis of unplanned data and the investi-
existence of a real difference. All it says is gator has to make assurances doubly sure
that the differences are very unlikely to have before drawing any conclusions. For this and
arisen by chance. Likewise, a ‘not significant’ other reasons, retrospective studies should not
decision does not ‘prove’ the absence of any be undertaken as an easy way out when
real difference ; it only means that the observed properly designed prospective studies are
difference could well arise by chance alone. possible and feasible.
Incidentally, it is also possible for a difference
to be statistically significant, and yet unimpor- After the completion of analysis, findings
tant. Provided a sufficiently large number of have to be presented either in the form of a
patients has been included in the trial, differ- paper in a journal or as a personal communi-
ence between sputum conversion rates of 70% cation to a scientific conference. The technique
and 72% in two drug schedules may well turn of writing research papers is not really difficult
out to be statistically significant. Yet the differ-to acquire. Papers, above all, should be read-
ence is so slight that it may not be worth able and simplicity should be aimed at, even if
bothering about and the medical worker may it needs a special effort. The question of the
well be justified in preferring the schedule with size and number of tables and charts is impor-
70% conversion if it has other advantages. tant, as we will presently discuss, but no less
important is that of a lucid exposition in the
Considering the stress Lald on proper plan- text. Indeed, the finest papers are those where
ning of studies, one might well get the impre- tables etc. do not impede the line of reasoning
ssion that studies on unplanned data are, at and can even be dispensed with for a prelimi-
best, a waste of time. This, however, is not minary reading.
always true. Retrospective studies necessarily
have to deal with unplanned data and may be It is difficult to lay down any rules regar-
inescapable in many cases. The analysis of Data ding the number of tables and diagrams but
from such studies however is a very complex too liberal a use of these certainly gives rise to
task. It is necessary to scrutinize the data in a sort of ‘consumer resistance’. There is also the
detail to eliminate possible source of bias. It is danger that with too many tables and charts-
not unusual that a finding which appears plausi- one may be unable to see the wood for the
ble on the face of it may turn out to be un- trees. In general, tables in the text should be
warranted on closer examination. Spurious brief and purposeful; the subsidiary ones are
associations in particular have to be guarded better relegated to the Appendix. Each table
against. An example may be worth quoting should be complete in itself, i.e., relevant
here. Some years back we at the New Delhi information such as abbreviations and units
TB Centre were interested in finding out whe- used, period covered, whether a certain rate is
ther the pattern of tuberculous disease had per cent or per thousand, or per thousand per
changed materially over the years. To this year, etc. should be given, within the table if
end, x-rays of a sample of cases reporting in possible, in footnotes if necessary. Graphs,
Ind. J Tub , Vol. XVI, No 2
ANALYSIS AND PRESENTATION OF DATA 53
bar diagrams etc. are sometimes very useful presentation has to be judiciously pruned.
for driving home a point but their use as a Only the most important tables and diagrams
routine or as a substitute for tabular data is to can be presented. These, again, should be as
be avoided. Serious minded readers sometimes simple as possible, each single slide containing
want to make some additional calculations on only the minimum quantity of figures needed
their own and only tabulated data rather than to make a point. It must be remembered that
sterile diagrams can serve this purpose. the human eye can take in only a limited
quantity of statistical data at a time— spe-
Having made a plea for brevity, it now cially when each slide is projected for just
remains, paradoxically, to enter a caveat about about 30 seconds. If too much mental effort
the dangers of going to the other extreme. is required to study a table or follow an
Instances are not rare of papers—and I mean argument many in the audience would rather
papers with a statistical aspect—published with give up. Since it is the speaker who is prima-
the sketchiest of data. Quite often, the idea at rily interested in ‘selling’ his paper, it is for
the back of the mind is that figures are boring him to forestall audience apathy.
things and ‘let us get them over while the fit is
on us’. This suggests a casual, almost frivo- I am quite conscious that it is no easy task
lous, attitude to the whole business of research for most clinicians to work up any enthusiasm
and only the uncritical can be taken in by such for figures and statistical methods. It will not
logic. The discerning reader is apt to ascribe be a bad beginning if these are accepted even
lack of relevant data to ignorance of the as a necessary evil. For, in the words of
research discipline, or worse. Bradford Hill, “In both clinical and preventive
medicine, and in much laboratory work, we
What has been said so far refers specially cannot escape from the conclusion that they
to papers meant for publication. Further (figures) are frequently cogent, that many of
problems arise when a paper has to be presen- the problems we wish to solve are statistical
ted in a conference instead of being printed in and that there is no way of dealing with them
a journal. As the time is limited, material for except by the statistical method”.
Ind. J. Tub., Vol. XVI, No. 2
AGAR ELECTROPHORESIS OF SERUM PROTEINS IN PULMONARY
K. L. AGRAWAL , S. NARASIMHA RAO and D. P. AGRAWAL
(From Kasturba Medical College, Mangalore.)
Introduction 5 ml. of blood was collected and allowed to
clot at room temperature. In most cases the
Reversal of serum albumin-globulin ratio sera were subjected for analysis, when it was
in a wide variety of diseases caught the fasci- not possible to analyse on the same day the
nation of research workers and gave them an sera were kept frozen till they were analysed.
impetus to penetrate into the intricate
mechanisms, if any, involved. Much break- Total serum proteins were estimated by the
through was made since a number of earlier biuret method of Kingsley (1942). Electro-
workers reported such a reversal in pulmonary phoresis was done according to the method
tuberculosis employing the conventional described by Giri (1956). The electrophoreto-
techniques of salt fractionation then in vogue gram so obtained was scanned by photoelectric
(Eichelberger & McCluskey, 1927 ; Gutman densitometer. A graph was constructed with
et at 1936 ; & Bing, 1940) The development the densitometric readings as the abscissa and
of electrophoretic technique of serum protein the distance in millimeter as the ordinate
fractionation by Tiselius in 1937 brought in a (Fig. 3 and 4). Alpha-1 globulin moved
revolutionary change in the concept of almost along with the albumin fraction, there-
albumin-globulin ratio reversal and afforded fore it was included in the albumin value.
the research workers a more meaningful tool The exact concentration of each protein
to understand the complex field of protein fraction was then obtained from the total
biosynthesis and protein alternations in health protein value.
and diseases. Luetscher (1941) was the first to
report electrophoretic studies in pulmonary E.S.R. for the first hour was estimated by
tuberculosis. He found decreased albumin, Westergren’s method. The presence of C-
increased alpha and gamma globulin levels in reactive proteins was qualitatively determined
his cases. By and large it appears that by slide-latex fixation test using latex anti C-
increase in alpha-2 globulin is the characteristic reactive protein reagent manufactured by
finding in pulmonary tuberculosis, although Hyland Laboratories, Los Angeles,
observations regarding alternations in the California, U.S.A. The test was performed as
other protein fractions are conflicting. The per directions provided with C-reactive
present study was undertaken to study the protein kit.
changes in serum proteins in pulmonary
tuberculosis employing agar electrophoretic Results
technique and to correlate, if possible, the
protein fractions with E.S.R. and C-reactive Electrophoretic serum protein values, E.S.R.
protein value. and C-reactive protein values of the 33 tuber-
culosis patients and 20 normal controls are
Materials and Methods presented in Table I—III. The mean total
protein value of 33 tuberculosis cases is
Thirty three patients attending the Wenlock 5.597±0.446 (S.E. Mean 0.079) and that of
Hospital, Mangalore, were selected for the the normal controls is 6.245±0.647 (S.E. mean
present study. The pulmonary lesions were 0.148). Statistical analysis of the values
subdivided into minimal (referred to as early revealed that the decrease in the total protein
by us) and moderately advanced (referred to as values in tuberculosis is highly significant with
advanced by us) in accordance with the a ‘t’ value of 4.225 (Table I). The observed
classification of National Tuberculosis Asso- decrease of the mean albumin value in the
ciation. All the patients came from the low tuberculosis patients as compared to the
middle class and lower classes of society. normal controls is not statistically significant.
Their nutritional status was low. No attempt Similarly the apparent increase in the average
has been made to group the cases according A/G ratio and the decrease of albumin/alpha-2
to age, or sex because of the small number of globulin ratio of the tuberculosis patients over
subjects. A group of twenty normal blood the corresponding average values for the
donors has been included to serve as control. normal group is not statistically significant.
However, the increase in alpha-2 globulin value
On the morning of the day of the test, and the decrease in beta and gamma globulin
Ind. J, Tub., Vol. XVI, No. 2
AGAR ELECTROPHORESIS OF SERUM PROTEINS IN PULMONARY TUBERCULOSIS 55
Mean and standard deviation of protein and electrophoretic values of 33 Tuberculosis
patients and 20 normal controls
Tuberculosis cases 20 Normal controls Difference in means of
patients and controls
Total Protein Mean 5.597 6.245 —0.648**
S.D. 0.446 0.647
S.E. 0.079 0.148 (t=4.225)
Albumin Mean 2.836 3.050 -0.214 N.S.
S.D. 0.661 0.237
S.E. 0.117 0.055
Alpha-2 Globulin Mean 0.963 0.580 0-383**
S.D. 0.415 0.415 (t=3.524)
S.E. 0.082 0.033
Beta-Globulin Mean 0.517 0.830 —0.3134**
S.D. 0.318 0.168 (t=4.727)
S.E. 0.056 0.037
Gamma-Globulin Mean 1.280 1.735 —0.455**
S.D. 0.517 0.504 (t=3.411)
S.E. 0.076 0.116
A/G Ratio Mean 1.167 0.977 0.190 N.S.
S.D. 0.626 0.165
S.E. 0.111 0.038
Albumin/Alpha-2 Mean 4.313 5.500 — 1.187 N,S.
ratio S.D. 3.626 0.987
S.E. 0.641 0.226
Note : N.S. — Not significant
** — Highly significant P 0.01
The t — Test for the difference of means is used in all these.
values of the tuberculosis patients compared to culosis group. No statistical significance has
the normal group is highly significant with ‘t’ been observed using the ‘t’ test.
values of 3.524, 4.727 and 3.411 respectively.
Statistical analysis of the E.S.R. and elec- Fig. 1 depicts the E.S.R. values of the T.B.
trophoretic protein fraction of the 25 advanced patients with positive and negative C-reactive
and 8 early cases have been presented in table protein. Fig. 2 indicates the total protein
II with a view to find out whether it is possible’ values of the T.B. patients with positive and
to assess the severity of disease as classified. negative C-reactive protein values. As eviden-
It can be seen from the table that except ced from the diagrams there is no cluster of
albumin/alpha-2 globulin ratio all other find- points at any place ; both the positive and
ings show statistically insignificant variation. negative values are scattered and hence statisti-
In the case of albumin/alpha-2 globulin ratio cally insignificant. Similarly no significant
early cases of tuberculosis have a mean value correlation could be obtained when positive
of 5.314±5.116 with a ‘t’ value of 2.767. This and negative C-reactive protein cases were
increase is highly significant. correlated with other electrophoretic protein
In table III are given the mean values of fractions. Figs. 3 and 4 show the typical
E.S.R. and electrophoretic protein fractions electrophoretograms obtained for a normal
for the 24 C-reactive protein positive and 9 and tuberculosis serum. The results have been
C-reactive protein negative cases of the tuber- summarised in table IV.
Ind, J. Tub., Vol. XVI, No. 2
56 K. L. AGRAWAL, S. NARASIMHA RAO AND D. P. AGRAWAL
Mean and standard deviation of E.S.R., total protein and electrophoretic values of Tuberculosis patients
Mean ± standard deviation of Difference in mean of
advanced and early
25 advanced 8 early cases ‘all 33 cases
E.S.R. (m.m. for 1st hour) 88.44 78.62 86.06 9.82 N.S.
±23.976 ±28.115 ±25.349
Total Protein (gm%) 5.652 5.425 5.597 0.227 N.S.
±0.433 ±0.444 ±0.446
Albumin 2.804 2.938 2.836 -0.134 N.S.
±0.653 ±0.675 ±0.661
Alpha-2 Globulin 0.965 0.955 0.963 0.010 N.S.
Beta Globulin 0.567 0.360 0.517 0.207 N.S.
±0.334 ±0.318 •e*
Gamma Globulin 1.3140 1.173 1.280 0.141 N.S.
±0.469 ±0.292 ±0.517
A/G ratio 0.105 1.361 1.167 -0.256 N.S.
±0.591 ±0.689 ±0.626
Albumin/A lpha-2 3.993 5.314 4.313 -1.321 **
ratio ±2.923 ±5.116 ±3.626 (t =2.767)
Note : N.S. — Not significant
** — Highly significant: P 0.01
Discussion It is believed that in most of the infectious
diseases there is a reversal of the albumin-
Intensive study of the distribution of serum globulin ratio. Our results in pulmonary
proteins in tuberculosis has received the tuberculosis do not subscribe to such a belief.
attention of many workers. Evidence has been Agar electrophoretic study of the serum protein
established that in “active infection”, signifi- changes in our series has shown that there is
cant changes in serum proteins occur. no statistically significant decrease of albumin
Gaitonde et al (1959) and Bovornkitti (1962) value in tuberculosis patients compared to
have reported a fall in total protein level in normal or between advanced and early cases.
pulmonary tuberculosis. We have observed
a similar fall in our cases of pulmonary tuber- The most striking finding of our results is
culosis. The decrease in our series of 33 cases the abnormality found in the globulin fractions.
when compared with normal group is highly Raised levels of alpha-2 globulin has been the
significant with a ‘t’ value of 4.225 (Table I). most consistent finding. This observation is in
This indicates poor nutritional status of tuber- confirmity with those of others (Gaitonde 1959,
culosis patients. The fall in total protein Bovornkitti 1962, Seibert 1942, Volk 1953, &
levels has been reported to be progressive from Gilliland 1956). The increase in alpha-2
minimal to advanced cases (Bovornkitti 1962). globulin fraction represents tissue destruction
However, this fall in our series is not progres- (Seibert 1947) and therefore it can be said that
sive since the value in early cases is 5.425 gs% pulmonary tuberculosis sets in tissue destruc-
whereas, in advanced cases it is 5.652 gs.%. tion in the host. This is so even in early
Ind. J. Tub., Vol. XVI, No. 2
AGAR ELECTROPHORESIS OF SERUM PROTEINS IN PULMONARY TUBERCULOSIS
Mean and standard deviation of the protein and electrophoretic values of T.B. cases with positive
and negative C-reactive protein
Characteristic Studied C-Reactive Protein Positive C-Reactive Protein Negative
(24 values) (9 values)
Mean Standard deviation Mean Standard deviation
E.S.R. (mm for 1st hour) 86.38 27.14 85.22 23.21
Total Protein 5.538 0.438 5.756 0.429
Albumin 2.729 0.576 3.122 0.777
Alpha-2 Globulin 0.981 0.467 0.916 0.454
Beta-Globulin 0.563 0.316 0.393 0.285
Gamma Globulin 1.267 0.471 1.314 0.309
A/G. ratio 1.068 0.540 1.432 0.750
Albumin/Alpha-2 ratio 3.948 3.041 5.287 4.745
For each of the characteristic studied above the difference between the mean of positives and of
negatives is found to be NOT SIGNIFICANT by using t — tests
There is no correlation between the C-reactive protein and any of the above characteristics.
Table showing summary of the results
Pulmonary T.B. total number of cases
compared with normal
Pulmonary T.B. early cases compared
with advanced cases
Pulmonary T.B. advanced cases
compared with early cases
Continuous and broken arrows show significant and insignificant direction of alteration respectively.
Ind, J. Tub., Vol. XVI, No. 2
58 K. L. AGRAWAL, S. NARASIMHA RAO AND D, P. AGRAWAL
Fig. I Fig. II
Fig. Ill Fig. IV
stages of the disease. The stage of the disease in the early stage of the lesion (table I). This
process, however, cannot be definitely assessed observation is not in confirmity with the find-
since there is no correlation in our series ings of other workers who observed a signi-
between the values and the severity of the ficant elevation of beta globulin fraction in
lesions. severely ill patients (Gaitonde 1959, Volk
1953, Seibert 1947). However, Baldwin and
Beta globulin values tend to fall in tuber- Hand (1953) have also observed a decrease in
culosis group as compared to the normal the beta globulin component. Since the increase
group. The decrease is significant with a ‘t’ of beta globulin is generally associated with
value of 4.727. The decrease is more marked liver damage, the differing observations may
Ind. J. Tub,, Vol. XVI, No. 2
AGAR ELECTROPHORESIS OF SERUM PROTEINS IN PULMONARY TUBERCULOSIS
not be ascribed primarily to pulmonary tuber- Summary
culosis. Further work is necessary to impli-
cate elevated beta glubulin as a secondary Fractionation of serum proteins by agar
effect of pulmonary tuberculosis or otherwise. electrophoresis has been done in 20 normal
There is significant fall of gamma globulin controls and 33 patients of pulmonary tuber-
fraction with a ‘t’ value of 3.411. This finding culosis prior to treatment with anti-tuberculosis
is not in agreement with other workers who drugs. The results clearly demonstrate that
have all reported an increase of this fraction in there are definite changes in serum proteins
all cases of active pulmonary tuberculosis. from normal values in tuberculosis. A signi--
Seibert et al (1942) have ascribed the rise to be ficant fall in total proteins, beta and gamma
an indication of resistance to the disease. It is globulins and a rise in alpha-2 globulin has
generally known that increased gamma been consistently observed. Albumin/alpha-2
globulin is a common characteristic of most globulin ratio in advanced cases has been
hepatic disease (Agarwal 1957, & Popper 1951) found to be significantly lower than that of
and that liver impairment is present in early cases.
advanced tuberculosis (Hurst 1947, & Small ACKNOWLEDGEMENT
1950). Accordingly the elevation found by
other workers could partly be expLalned by The authors are thankful to Dr. M.V.
liver alteration. Since our series did not Chari, Principal, Kasturba Medical College
include far advanced cases, it is likely that for permission to publish this paper and to
they did not have associated liver alteration. the Department of Biostatistics, Christian
Since our series did not include far advanced Medical College, Vellore for statistical analyses.
cases, it is likely that they did not have
associated liver alteration. However, the REFERENCES
significantly low value obtained by us remains 1. Agarwal, K.L., Kumar, A. Kumar, S,, and
to be expLalned. Mangalik, V. S. (1957) J. Ind. Med. Assoc., 29
The clinical status is better correlated with (10), 387.
albumin/alpha-2 globulin ratio and it is con- 2. Baldwin, R.W.. and Hand, C.N. (1953) Amer.
sidered to be a better criterion of the severity Rev. Tuberc. 68, 372.
of the disease process. We have also observed 3. Bing, J., (1940) Acta. Med. Scandinav, 103, 547.
albumin/alpha-2 globulin ratio fall in tuber- 4. Bovornkitti, S., (1962) Amer. Rev. Resp.
culosis when considered as a whole. But this Diseases, 85, 58.
alteration is not statistically significant. How- 5. Eichelberger, L., and McCluskey, K.L. : (1927)
ever, there is a definite and significant fall of Arch. Int. Med., 40, 831.
this ratio in the advanced stage of the disease 6. Gaitonde, B.B., Shinde, A.G. and Rao G.S. :
(Table I). Thus our observation is in confir- (1959) Ind. J. Tub. 6, No. 1, 12.
mity that the albumin/alpha-2 globulin ratio 7. Gilliland, I.G., Johnstone, R.N., Stradling,
is a better criterion of the severity of the
disease process as observed by other workers. P. and Abdelwaheb. E.N. : (1956) Brit. Med. J.
That E.S.R. value cannot be correlated
with the protein fractions and the severity of 8. Giri, K. V. : (1956) J. Lab. Clin. Med, 48,
the disease is evident from Table II. Our No. 5, 775-778.
findings are in agreement with Gaitonde 9. Gutman, A.B., Gulman, E.B., Jillson, R., and
(1959), & Gilliland (1956) that correlation of Williams. R.D. (1936) J. Clin. Investigation, 15,
the clinical status with albumin/alpha-2 475.
globulin ratio is a better index than with 10. Hurst, A., Maier, H.N., and Lough, S.A. : (1947)
E.S.R. Amer. H. Med. Sci. : 214, 431.
Although biochemical study of the alter- 11. Kingsley : (1942) J. Lab. Clin. Med., 27, 840.
ation of the serum protein fractions in pulmo- Luetscher, J.A. Jr., (1941) J. Clin. Invest., 20,
nary tuberculosis alone do not prove to be of 99.
diagnostic value, they will fruitfully assist in 12. Popper, H., Bean, W.B., De La Huerga, J.,
establishing the therapeutic efficacy and the Franklin, M., Tsumagari, Y., Routh. T.I. and
Stiegmann. F., (1951) Gastroenterology, 17, 138.
severity of the lesion. “The protein spectrum
in plasma and serum is the resultant of a host 13. Seibert, F.B., Seibert, M.V., Atno, A.J. and
Campbell. H.W. : (1947) J. Clin. Invest., 26, 90.
of factors concerned with the formation, the 14. Seibert, F.B., Nelson, J.W., (1942) J. Biol.
interaction, and the destruction of the indivi- Chem. : 143, 29.
dual components”. It remains, however, to 15. Small, M.J. : (1950) Amer. Rev. Tuberc. 61, 893.
be established what interplay of mechanisms Teselius, A., (1937) Biochem J., 31, 1464.
are responsible for the alterations in the 16. Volk, B.M., Salfer, A., Johnson, L.E., and
different globulin fractions. Oreskes, I. (1953) Amer. Rev. Tuberc., 67, 299.
Ind. J. Tub., Vol. XVI, No 2
ETHIONAMIDE AND ISONIAZID IN THE MIDDLE-AGED
AND THE ELDERLY PATIENTS
R. K. NARANG
(From G. S. V. M. Medical College, Kanpur)
Introduction were included in the study. Moribund patients
were excluded from the trial.
A combination of streptomycin and isonia-
zid is very effective in the initial treatment of Sputum was examined for AFB using the
pulmonary tuberculosis. However, strepto- Ziel-Neelson technique. Only smear positive
mycin, since it is given by injection, has cases were included. As the culture facilities
obvious drawbacks in ambulatory therapy. in the department are inadequate and unreli-
There are villages where the patients have to able, no culture examination was done. Smear
walk miles to get their daily injection. The examination was repeated once a month and,
charge for injection adds to the cost of treat- if found negative, further specimens were
ment. Again, streptomycin is not a very suit- examined on three consecutive days. If no
able drug for the middle aged and the elderly AFB were demonstrable in all the specimens,
patients due to high incidence of vestibular and the case was adjudged as bacteriologically
cochlear toxicity. negative. If bacteriological conversion was
not obtained at four months and especially
Encouraged by the report of British Tuber- if there was also evidence of clinical or radio-
culosis Association/Hong Kong Tuberculosis logical deterioration, the case was withdrawn
Treatment Services (1964) on ethionamide and from the study, having been adjuged as treat-
isoniazid in newly diagnosed previously un- ment failure.
treated cases of pulmonary tuberculosis, it was
decided to assess the role of this combination A chest radiograph was taken in the begin-
in the middle-aged and the elderly cases of ning in every case to assess the extent and
pulmonary tuberculosis. type of the disease. Only moderately advanced
or far advanced cases were selected for the
Material And Methods trial ; minimal cases were excluded as excellent
results have been reported with isoniazid alone
The material for the present study con- in such cases. The skiagram was repeated
sisted of 50 cases of pulmonary tuberculosis every three months and at the time of sputum
whose ages ranged from 50 to 85 years conversion.
(average 62 years). 44 were men and 6
women. 2 cases were, however, withdrawn All the patients were unhospitalized
from the trial as they were found to have had throughout the course of their treatment.
previous treatment with anti-tubercular drugs. They collected drugs for one week in the first
Another 3 cases were withdrawn early in the instance and if no side effects to the drugs were
trial as they did not stick to the protocol ; a compLalned of, they were supplied drugs for a
totally oral regime did not appeal to them and fortnight at a time.
they were getting streptomycin injection from
other sources. On each visit the progress was assessed
clinically ; radiographs and sputum examina-
All the cases were placed on 400 mg. tions were ordered as outlined above. Any
isoniazid and 500 mg. of ethionamide daily symptoms or signs suggestive of drug toxicity
given as a single dose at bed time. Isoniazid were noted ; to avoid suggestion no attempt
was given as four 100 mg. tablets and ethion- was made to elicit symptoms by direct ques-
amide as four 125 mg. sugar coated tablets. tioning.
Although higher strength ethionamide tablets
are also available, 4 tablets of each drug was Observations And Results
thought to lessen the chances of misunder-
standing about instructions. It was proposed As mentioned above 5 cases were excluded
to continue the combination for one year. The early in the trial due to various reasons. Thus
drugs were supplied free of cost to the only 45 cases were assessed. Of these 23 had
patients. moderately advanced disease and 22 had far
advanced lesion. 12 cases also suffered from
The patients were closely questioned about chronic dyspepsia, 3 had diabetes mellitus (1
the history of previous treatment, if any. Only with neuropathy and nephropathy also) and 4
those who had not received any specific therapy had history suggestive of chronic bronchitis.
Ind. J. Tub.. Vol. XVI, No. 2
ETHIONAMIDE AND ISONIAZID IN THE MIDDLE-AGED AND THE ELDERLY PATIENTS 61
A. Efficacy of Treatment (Table Nos. 1 & 2) TABLE 1
Of the 45 cases, 2 cases had to be with- Showing the results of treatment
drawn due to intolerance to the drugs (pre- Total number of cases—43
sumably to ethionamide), so that efficacy No. of cases Percentage
of treatment was assessed in 43 cases only.
Clinical improvement, as shown by relief Clinical improvement 42 97.6
in chest symptoms and toxemic manifestations
were noted in 42 out of 43 cases. 3 of these Radiological improvement 42 97.6
Bacteriological conversion 39 90.0
All the cases had cavitary disease on
admission to the trial. In 97.6% there was
radiographic improvement, assessed by clearing TABLE 2
of shadows and/or dimunition in size of the
cavities. In 31 cases (72%) cavities dis- Showing the bacteriological response in relation
to extent of disease
appeared completely while in others they were
markedly reduced in size. The maximum Total number of cases—43
improvement was noted in the first three Extent of disease No of No. of Percent-
months although further radiological clearing
continued to occur upto nine months. cases cases age
Sputum converted in 39 cases (90%). In
25 this occurred in first 2 months, in 10 in the Moderately advanced 21 19 90.0
third month and in 4 in the four month. The
speed and incidence of sputum conversion was Far advanced 22 20 90.0
the same in moderately advanced and in far
advanced cases. All the 39 cases maintained
the bacteriological improvement, 35 cases com- TABLE 3
pleted treatment for full one year, 3 cases for Showing the adverse reactions to drugs
9 months and one for 8 months. The four
cases who responded to treatment but did not Total number of cases-43
complete full one year treatment did not report Nature No. of Percent-
for collection of drugs and could not be cases age
Drug Toxicity and side effects (Table No. 3) Gastrointestinal
Adverse reactions to the drugs were com- Persistent 2 4.4
mon but were not serious enough to need with- Minor 7 15.5
drawal of the drugs except in two cases.
Neuropathy (Subjective) 2 4:4
A metallic taste in the mouth, anorexia,
nausea or vomiting was compLalned of by 9 Allergic 2 4.4
cases (19.9%). In 2 cases the symptoms were Miscellaneous 6 13.3
persistent and intractable and the patients
refused to continue the treatment. In the other
7 the symptoms became less or disappeared in
course of time. allergy to one of the drugs. However, these
symptoms could be controlled by small doses
Burning sensation in the palms and soles of anti-histaminics.
and pain in the limb were compLalned by 2 3 cases compLalned of drowsiness and
cases (4.4%). These symptoms were controlled lethergy ; one of them was a diabetic. Another
with 10 mg. pyridoxine hydrochloride per day 2 cases compLalned of heaviness in the head.
while the anti-tubercular drugs were continued. Gynecomastica with acne was noted in one
In no case there was objective evidence of case. These symptoms needed re-assurance
In 2 cases (4.4%) a transient rash and No case of jaundice or exfoliative dermati-
itching occurred which was attributed to tis was noted in the present series.
Ind. J. Tub.. Vol. XVI, No. 2
62 R. K. NARANG
Discussion sputum conversion in 77% of 13 cases at 6
In the present study 400 mg. isoniazid and Gastrointestinal side effects were noted in
500 mg. ethionamide was given orally in a 20% cases in the present report but except in
single dose at bed time in middle aged and two cases they were not troublesome enough
elderly patients of pulmonary tuberculosis who to need withdrawal of the drugs. Thus in
were not previously treated with anti-tuber- 500 mg. dose daily ethionamide was very well
cular drugs. The dose of ethionamide is tolerated. Good tolerance was also observed
smaller than used in the British Tuberculosis in other series mentioned above. This is in
Association (1961) investigation. There is a marked contrast to the high incidence of
good experimental support for using small intolerance in the British Tuberculosis Associa-
doses of ethionamide with isoniazid. Thus tion trial (1961) of a combination of ethiona-
Mm. Grumbach (1963) concluding on the mide, pyrazinamide and cycloserine. In a
basis of experimental tuberculosis in mice more recent study by the British Tuberculosis
recommends 400 to 500 mg. isoniazid (corres- Association (1968J, again a high incidence of
ponding to 4 times the minimal effective dose) gastrointestinal symptoms were reported. In
with 200 to 500 ethionamide. Again RIST this study ethionamide was combined with
(1964) as a result of laboratory experiments streptomycin and isoniazid. The dose of
observed that in isoniazid-ethionamide combi- ethionamide in both the trials was high. That
nation isoniazid is the principal drug and the dose of ethionamide is one of the impor-
should be given in high doses while 500 mg. tant factors in relation to the incidence of
ethionamide is adequate. gastrointestinal side effects is also suggested by
a trial of ethionamide, isoniazid and thiaceta-
Initially 50 cases were selected for the trial. zone in drug resistant cases by the present
However, 5 cases did not satisfy the protocol author (Narang and Sarin, 1966). In this
and were withdrawn. Another 2 cases were investigation 750 ethionamide in two divided
withdrawn in the first fortnight due to gastro- doses was used and a high incidence of gastric
intestinal intolerance to the drugs. Thus upsets was observed. Other possibly important
efficacy of the drugs has been analysed in 43 factors are the drug combinations and racial
cases and side effects to the drugs in 45 cases. differences.
Sputum conversion was obtained in 39 of Allergic manifestations were uncommon
the 43 cases (Conversion rate 90%) who and could have been due to ethionamide or
remained in the trial sufficiently long. If all isoniazid. Carey (1965) reports successful
the 45 cases are included, the conversion rate desenstisation in a case of ethionamide allergy.
Minor side effects included drowsiness,
There are three other reports in the lite- lethargy, heaviness in the head, gynecomastica
rature on the use of ethionamide-isoniazid and acne. Drowsiness may be a manifestation
combination in previously untreated cases. of hypoglycemia (British Tuberculosis Associa-
Our material, in [contrast to these reports, tion, 1968). Hypoglycemia has been described
consisted of middle aged and elderly patients in diabetic patients treated with ethionamide
only. Again, while in the present report the (Clark and O’Hea. 1961, Somner and Brace,
doses of ethionamide and isoniazid were 1962). Of the three patients in the present
500 mg. and 400 mg. respectively given in a series, who compLalned of drowsiness, one
single dose, Lees (1964) from Glasgow used was a diabetic. But investigations to exclude
1 gm ethionamide with 400 mg. isoniazid, hypoglycemia were not done.
British Tuberculosis Association/Hong Kong
Tuberculosis Treatment Services (1964) and Clinical jaundice was not noted in a single
Bhatia and Lal (1966) from Amritsar used case. This is especially striking considering
500 mg. ethionamide and 300 mg. isoniazid. that the 3 of the cases in the present study
While in the Hong Kong trial the drugs were were known diabetics ; the diabetics and the
given in a single dose, the Amritsar workers alcoholics are especially prone to develop
gave the drugs in two divided doses. hepatotoxicity (De Voogd, 1963). The low
incidence of icterus has also been reported by
In the Hong Kong trial, sputum conversion other workers. However, routine liver func-
was obtained in 98% of cases who continued tion tests are likely to reveal abnormality in a
the drug for one year, overall conversion rate higher percentage (Bhatia and Lal, 1966 ;
was 85% including those who left off. Lees Somner and Brace, 1967 ; British Tuberculosis
(1964) obtained sputum conversion in all the Association, 1968). No routine liver function
32 cases. Bhatia and Lal (1966) noted a tests were done in the present study.
Ind. J. Tub , Vol. XVI, No. 2
ETHIONAMIDE AND ISONIAZID IN THE MIDDLE-AGED AND THE ELDERLY PATIENTS 63
Some of the treatment failure cases might REFERENCES
have been primarily resistant to one of the
drugs. As sensitivity tests were not done, their 1. British Tuberculosis Association ( 1961 ):
incidence is not known. There is a very Tubercle, Land. 42, 269.
interesting relationship between isoniazid
resistance and ethionamide. While low degree 2. British Tuberculosis Association/Hong Kong
isoniazid resistant bacilli are sensitive to Tuberculosis Services (1964) : ibid, 45, 299.
ethionamide, in man there is usually a high
degree isoniazid resistance even in rapid 3. British Tuberculosis Association (1968) : ibid.
inactivators so that ethionamide is likely to be 49, 125.
effective in such cases. Canneti (1965) and 4. Bhatia, J. L. and Lal, H. (1966) : Indian J.
Nazaki (1964) in experiments using H37Rv Tuberc. 13, 57.
strain of mycobacterium tuberculosis resistant
to 50 meg per cc. of isoniazid found that the 5. Canneti, G. (1965) : Amer. Rev. Resp. Dis.
combination of ethionamide and isoniazid had 92, 687.
higher antibacterial activity against the test
strain than ethionamide alone and observed 6. Carey, V.C.I. (1965) : Tubercle, Land., 46, 287.
decreased resistance of resistant strain to
isoniazid. They confirmed these findings in 7. Clark, G.B.M. and O’Hea, A.J. (1961) : Brit.
male mice. Med. J., 1, 636.
Summary 8. De Voogd, A. (1963) : Rev. Tuberc. 27, 935.
Abstracted Amer. Rev. Resp. Dis. 91, 158, 1965.
The role of 500 mg ethionamide and
400 mg isoniazid, given in a single dose orally 9. Mm. Grumbach, F. (1963): Abstr. ibid, 87, 800.
at bed time was assessed in 45 previously un- 10. Lees, A.W. (1964) : Dis. Chest. 45, 247.
treated middle aged and elderly cases of pul-
monary tuberculosis. 90% sputum conversion 11. Narang, R.K. and Sarin, J.N. (1966) : J. Assoc.
was obtained. If those who fell out of Phys. India, 14, 503.
the trial due to drug intolerance the conversion
rate was 81.5%. The drugs, in the dosage 12. Nazaki, T. (1964) : Kekkaku, 39, 27, Abstr.
given, were very well tolerated. Amer. Rev. Resp. Di. 91, 808, 1965.
ACKNOWLEDGEMENT 13. Rist, N. (1964), Selected Papers of Royal
Netherlands Tuberculosis Association, 8, 24.
The author acknowledges the generous
supply of ethionamide by M/s. Themis 14. Somner, A.R. and Brace, A.A. (1967), Tuberc.
Pharmaceuticals. Lond.,49, 137.
Ind. J. Tub., Vol. XVI, No. 2
A COMPARATIVE CLINICAL EVALUATION OF THE ROLE OF THIOACETAZONE
AND PAS IN THE MANAGEMENT OF PULMONARY TUBERCULOSIS
B. K. KHANNA
(From K. G’s Medical College, Lucknow)
Introduction 2.4.2. According to previous treatment
the distribution of cases was as follows
Thioacetazone has, during the last few years, (vide table No. 2).
raised many points of controversy. While its
efficacy as tuberculostatic agent has been 2.4.3. According to various treatment
compared with that of PAS, its toxicity in vivo schedules, the distribution of cases was as
studies has been found to be variable (Miller follows (vide table No. 3).
Fox and Tall, 1966, Aquinas 1968 and Miller,
1968). It has been suggested that latter TABLE 1
manifestations may vary with the dietary Final distribution of cases
habits (Miller, Fox and Tall, 1966) and race. PAS Thioacetazoue
(Miller, Fox and Tall, 1966 and Aquinas, Group Group
We have been using Thioacetazone in Total number of cases
(Initial) 207 193
various combinations in our hospital for the Cases dropped from trial 4 6
past 5 years, and have conducted a controlled
trial on the relative efficacy of thioacetazone Reasons: Incomplete data 1 2
and PAS in the management of cases suffering
from pulmonary tuberculosis. The present Incorrect History 1 1
report pertains to that aspect of the problem. Left against advice 2 3
Material & Methods
Cases available for final
390 cases suffering from pulmonary tuber- analysis 203 187
culosis, admitted to Kasturba Hospital, were
studied. These cases were divided by. random TABLE 2
sampling into two groups—one group was Distribution of cases according to previous
recepient of PAS in conjunction with other chemotherapy received by the cases.
anti-tuberculosis drugs while the other received PAS Group Thioacetazone Group
thioacetazone in place of PAS. Exact dosage
and drugs schedules are given below: Untreated 108 91
All the cases at the initiation of trial had to Treated 95 96
fulfil following criteria:
2.1. Sputum must be positive for AFB Total 203 187
2.2. Radiological picture should be com TABLE 3
patible with the diagnosis of pulmonary Distribution of cases according to various
tuberculosis with atleast one cavity in schedules of treatment.
either infraclavicular region. PAS Group Thioacetazone Group
2.3. Every case was asked in detail
regarding the amount of drug treatment he SPH PH STH TH
had received in past. Those who had taken
less than 10 days of chemotherapeutic Untreated 52 56 42 49
treatment prior to their admission to the
hospital were labelled as ‘untreated’ and Treated 61 34 57 39
those who had more than 10 days’ therapy
were labelled as ‘treated’.
Total 113 90 99 88
2.4.1. The final distribution of cases was
as follows (vide table No. 1).
Ind. J. Tub., Vol. XVI, No. 2
ROLE OF THIOACETAZONE AND PAS IN PULMONARY TUBERCULOSIS 65
Schedule of Treatment Chest X-ray was repeated every three
S+T+H—Streptomycin IG IMI once a
day—Thioacetazone 150 mgms/day—INH 10 2.7. Assessment of results were based
mgms/kg. body weight—vitamins (hitherto emperically as under:
referred to as STH group).
(a) Improved: Sputum conversion.
T+H—Thioacetazone and isoniazid admi-
nistered in the same dose as above (hitherto Sub-Group: Marked Improvement—Sputum
referred to as TH group). conversion with disappearance of cavity.
S+P+H—Streptomycin and Isoniazid Moderate: Sputum conversion with
administered in usual dosage—PAS 10 G/day reduction in size of cavity.
(hitherto referred to as SPH group).
Mild: Sputum conversion with cavity
P+H—PAS and INH administered in same represented by a bullous cyst.
dose as in SPH group (hitherto referred to as
PH group). (b) Stationery: No sputum conversion, chest
After 3 months of daily injections of
Streptomycin the injections were withdrawn (c) Deteriorated: Deterioration on radiological
and patients received only TH or PH group of and clinical grounds or development of
drugs. toxic reactions to the chemotherapy.
2.5. Duration of Treatment Results
Duration of therapy was 6 months in each Final overall results have been depicted in
case. All the cases had to stay in the hospital Tables No. 4, 5 and 6.
during the period of trial. 5 cases who had
left the hospital prematurely had to be with- Discussion
drawn from the present trial (vide Table
No. 1). This study relates to hospitalised cases and
extends over a period of six months’ observa-
2.6. Investigations tion.
Sputum examination for AFB (Smear In the hospital itself drug administration
examination by ZN technique) was done every was done by the trained nurses who would
month. Gaffkey count was taken as a guide to ensure regular intake of adequate dosage of
progress. drugs in their presence. While Thioacetazone
Overall results at the end of six months’ study.
SPH (113) STH (99) PH (90) TH (88)
Improved 72(64%) 60(61%) 56(62%) 46(52%)
Marked: 25(22%) 12(12%) 11(12%) 10(11%)
Moderate: 42(37.5%) 38(39%) 26(29%) 13(15%)
Mild 5(4.5%) 10(10%) 19(21%) 23(26%)
Stationary 22(18%) (15(15%) 1102%) 9(10%)
Deteriorated 16(14%) 19(19%) 21(23%) 28(32%)
Expired 2( 3%) 2( 2%) 2( 3%) 1(1.5%)
Therapy changed U 1%) 3( 3%) — 4(4.5%)
Ind. J. Tub., Vol. XVI, No.. 2
66 B. K. KHANNA
Ind. J. Tub., Vol. XVI, No. 2
ROLE OF THIOACETAZONE AND PAS IN PULMONARY TUBERCULOSIS
SPH STH PH TH
(113) (99) (90) (88)
1. Ototoxicity 3 8 — 2
2. Cutaneaus Rashes — 5 — 3
3. Cutaneous rashes severe
enough to warrant a
withdrawal of drugs. — 2 — 1
4. Hepatic toxicity — 3 1 3
5. G.I. Intolerance 2 5 3 8
6. Peripheral Neuritis 1 2 1 1
1. One case in STH group died due to severe hepatic damage
passing on to hepatic coma.
2. One case in STH group and one in TH group developed such a
severe cutaneous reactions that withdrawal of therapy and
administration of corticosteroids had to be resorted to, in an
attempt to save their lives.
and isoniazid tablets were swallowed in a viomycin, cycloserine, pyrazinamide, ethiona-
single dose, PAS and INH were given in 2 mide, kenamycin, ethambutal etc.). Further
divided doses administered after meals. Thus, tolerance to these drugs is very poor. There-
the problem of drug default was almost com- fore, irrespective of the drug resistance pattern
pletely eliminated. However, since the bulk of of the tubercle bacilli excreted by the patient,
PAS and Isoniazid was quite different from an attempt was made to evaluate the efficacy
that of Thioacetazone and Isoniazid, it was of Thioacetazone in combination with Isoniazid
not possible to perform a double blind trial on alone or in conjunction with Streptomycin.
the subject. The results in the irregularly treated group of
the cases are disappointing and are certainly
A glance through table Nos. 4, 5 and 6 not better than PH or SPH groups.
reveals certain outstanding features. These can
be summarised as under: 4.2. The tolerance to thioacetazone was
poorer as compared to that of PAS. Both
4.1. Clinical response to Isoniazid and major and minor toxic reactions were met with
Thioacetazone combination was poorer as much more frequently in the former series.
compared to that of Isoniazid and PAS group.
This was especially true for those cases who 4.2.1. Jaundice appeared during thioaceta-
had not received previous treatment with anti- zone therapy in six cases (2 cases in untreated
tuberculosis drugs. group and 4 in treated group) as compared to
that of one (belonging to untreated group)
Culture and sensitivity studies on tubercle case in PH group. In one case of the former
bacilli were not done although the facilities for group, the patient passed on to hepatic coma
the same did exist in our hospital. During the and could not be saved. However, in remain-
trial our attempts were to create condition ing five cases belonging to TH group thioace-
which can be reproduced in any hospital. Most tazone in test dosage (1 mgm.) was again
of the cases getting admitted to our hospital started after the remission of jaundice. There
are those who have been grossly (and irregu- was no recurrence of jaundice. Thereafter, full
larly) treated in the past without deriving any dose of thioacetazone was instituted. This led
benefit. They have meagre resources at their us to conclude that jaundice in these cases
command and may not be able to afford could have been due to some other cause (e.g
second line of anti-tuberculosis drugs (e.g. infective hepatitis or homologous serum
Ind. J. Tub., Vol. XVI, No. 2
B. K. KHANNA
jaundice). However, its predominance in the study on thioacetazone (Miller et al, 1966) has
thioacetazone group might be regarded to already emphasised these points and has
indicate a casual relationship. It is also likely further emphasised that before embarking on a
that hepatotoxicity due to thioacetazone might mass scale use of this drug on a domiciliary
be due to cumulative toxic reactions of the basis, its tolerance in that particular commu-
drug and it might not have re-manifested itself nity must be ascertained. Our study has
during the remaining period of hospitalisation demonstrated that thioacetazone is not so well
of the patients. tolerated as PAS atleast by the residents of
Utter Pradesh, and that it is liable to lead to
4.2.2. Severe expholiative dermatitis was serious toxic reactions which may occasionally
met with in 3 cases in the thioacetazone group be fatal.
(all belonging to ‘treated’ group). In two cases
withdrawal of drug and symptomatic therapy Summary
failed to produce any improvement. 60 mgms
Prednisolone per day had to be instituted to 390 cases suffering from pulmonary tuber-
save their lives. These cases were later with culosis admitted to Kasturba Tuberculosis
drawn from the group. Reinstitution of isoni- Hospital were divided into four groups by
azid and streptomycin was not followed by random sampling. 203 cases received PAS
any cutaneous reaction in this group of cases. (10 gm/day) and isoniazid (400 mgms/day), 113
However, we avoided further use of thioace received streptomycin (1 G IMI7 once a day)
tazone in them. for first 3 months of observation in addition.
187 cases received Thioacetazone (150 mgms/
4.2.3. Thioacetazone group of cases suffe- day) instead of PAS along with isoniazid (in
red from toxic reactions to streptomycin on 8th usual dosage), 99 received streptomycin also.
nerve more frequently as compared to that of Our study extended over a period of six
PAS group. months and all along these cases were hospi-
talised. It was noted that the response to
Potentiation of ototoxic reaction of strep- thioacetazone was inferior to that of PAS
tomycin by thioacetazone (not by isoniazid, particularly in those cases who had not recei-
because it was a common drug in both the ved chemotherapy in past. Further, the
groups) permits us to conclude that combina- tolerance to thioacetazone was found to be
tion of thioacetazone and streptomycin should poorer than that of PAS.
be avoided in presence of obvious renal damage
and in persons beyond the age of 40 years. ACKNOWLEDGEMENTS
Of 8 cases in the STH group who had developed
this toxicity 5 were beyond 40 years of age. The author is grateful to Drs. J. Nath and
Two cases receiving thioacetazone and isonia- S.P. Misra for their cooperation in the trial.
zid alone developed ototoxic reactions (vesti- His grateful thanks are also due to M/s.
bular damage). This implies that thioacetazone Unichem Laboratories for the liberal donations
besides potentiating the vestibular damage due of Unithiben VF tablets containing Thioace-
to streptomycin, could by itself damage the tazone, isoniazid, antihistamine and vitamins.
vestibular system independently: although the
number of cases is too small to arrive at a REFERENCES
definite conclusion (Deshmukh and Master,
1962, Miller, Fox and Tall, 1966 and Miller, 1. Aquinas, M. (1968) Side Effects and Toxicity to
Thioacetazone and Isoniazid, Finding in A
1968). Hongkong Tuberculosis Treatment Service/
British Medical Research Council Investigation.
The study permits us to conclude that on Tubercle (Land.), Supplement to Vol. 49; 56.
clinical grounds thioacetazone is inferior to
that of PAS and is more toxic than PAS (with 2. Deshmukh, M.D. and Master, T.B. (1962)
Thioacetazone and Isoniazid in the Treatment
reference to both minor and major toxic of Pulmonary Tuberculosis J. Indian Med., Prof.
reactions). 9, 4273.
This observation, however, may not apply 3. Miller, A.B., Fox, W. and Tall, R. (1966), An
universally in all the cases of pulmonary tuber- International Cooperative Investigation into
Thioacetazone (Thioacetazone) side effects.
culosis because it has been demonstrated that Tubercle (Land.), 47; 33.
tolerance to thioacetazone may depend on race
(Aquinas, 1969), nutritional status of the host 4. Miller, A.B. Thioacetazone Toxicity: A General
and many other factors. The international Review. Tubercle (Land.), Supplement 49; 54.
Ind. J, Tub., Vol. XVI, No. 2
NEWS & NOTE
Annual Meetings Shri B.M. Cariappa, the Secretary-General,
also visited Goa and attended a meeting of the
The Thirteeth Annual General Meeting of Executive Committee of Goa Association and
the Tuberculosis Association of India was held addressed the Rotarians and Care-Committees
on 18th April at 11.30 A.M. in the Conference in Panjim.
Hall of the Association. Dr. S. Chandra-sekhar,
the President of the Association, presided. The West Bengal Conference
Chairman of the Association presented the
report on the working of the Association The Bengal Tuberculosis Association held
during 1968 and the Honorary Treasurer the 2nd West Bengal TB Conference from 12th
presented the accounts. The meeting elected to 14th April, 1969 in Calcutta. The confe-
members to the Central Committee as provided rence included three symposia, and a few origi-
for in the rules. nal papers were also presented.
The Conference of the Secretaries of the On this occasion, the Association brought
State TB Associations and Seal Sale Organisa- out an attractive Souvenir containing useful
stions in India was held in the Conference Hall information on the development of tuberculosis
of Association at 3.00 P.M. on 18th April and services in the State. The Association also
the Technical Committee of the Association inaugurated at this time a special chest
meet on 19th April. clinic for children, the first of its kind in India
containing advanced laboratory facilities for 42
VII Maharashtra Conference affiliated chest clinics. The three-storeyed
building was built at a cost of Rs. 2,25,000.
The Maharashtra State Anti-TB
Association organised a two-day State TB and Conference in Assam
Chest Diseases Workers Conference in
Bombay from 22nd to 24th March, 1969. The The TB Association of Assam will be hold-
Conference which was inaugurated by Dr. P.V. ing a Seminar on Tuberculosis shortly. The
Cherian, the Governor of Maharashtra was Seminar is likely to be attended by Dr. P.K.
addressed by Smt. Pratibha D. Patil, Deputy Duraiswamy, Director General of Health
Minister for Public Health and Prohibition. Dr. Services, and Shri B.M. Cariappa, Secretary-
P.K. Duraiswamy, Director General of Health General, Tuberculosis Association of India.
Services and Chairman of the Tuberculosis Award to Dr. R. Viswanathan
Association of India, also addressed the confe-
rence. Dr. R. Viswanathan, Emeritus Scientist,
Vallabhbhai Patel Chest Institute, University
Shri B.M. Cariappa, Secretary-General, of Delhi, Delhi, has been awarded the Eugenic
Tuberculosis Associations of India, was the Morathe Prize by Academia De Lincie of Italy,
President of the Conference. Shri Cariappa’s for outstanding contribution in the field of
address covered the role of voluntary TB Asso- Tuberculosis and Chest Diseases. The prize
ciations in India with special reference to has been awarded only to three other distingui-
Maharashtra. shed scientists so far in the world. Dr.
Viswanathan is the first non-Italian to receive
Discussions included ‘Control of Tuberculo- the Award, the value of which is two million
sis in rural areas’ in which Dr. B.B. Yodh and Liras.
P.A. Deshmukh participated. Dr. M.D. Desh-
mukh and Dr. J.C. Kothari took part in the Seal Sale Award—1969
discussion on ‘Place of Thiacetazone in treat- The 1969 Trophy for the highest Seal Sale
ment of Pulmonary Tuberculosis’. In another collections, was awarded to the Tamilnadu
discussion on ‘Tuberculosis in Children’ Dr. TB Association. The Trophy was presented
M.M. Wagle and Dr. M. Asher participated. to the Association at the time of the Annual
General Meeting of the Association held on
The Conference concluded with a Meeting 18th April, 1969.
of the Secretaries of District Associations in
Maharashtra. Proceedings of the XXIII TB Conference
On the occasion of the conference the Copies of the Proceedings of the Twenty-
Maharashtra Association brought out an inte- third Conference held in Bombay in 1968 are
resting handbook on Tuberculosis.
Ind. J. Tub., Vol. XVI, No. 2
70 NEWS & NOTE
available for sale from the Tuberculosis Asso- in different disciplines of Medical sciences on
ciation of India, 3-Red Cross Road, New an All India basis with a view to admit candi-
Delhi-1. The price per copy is Rs. 23/- plus dates to the Membership of the Academy. The
postage. next examination will be held in July, 1969 in
Delhi only. Application forms can be obtained
Chest Diseases’ Prize Award from the Executive Director, Indian Academy
The Indian Association for Chest Diseases of Medical Sciences, C-II/2, Medical Institute
has instituted a cash prize of Rs. 200 to be Campus, Ansari Nagar, New Delhi-16.
given to the author of the best article published
during the previous year either in Indian or In another announcement, the Academy
foreign Journal on any subject in the speciality has also invited young scientists engaged in bio-
of chest diseases. The prize is open to only medical research to participate in the Scientific
those who are under the age of 40 years. The session of its annual meeting to be held in
work on which the article is based must have
been conducted in India. Details can be had December, 1969 and is open to scientists of the
from the Secretary, Indian Association for age of forty and below. Selected scientists
Chest Diseases, Silver Jubilee TB Hospital, will be paid travelling and daily allowances for
Kingsway, Delhi-9 by 31.7.1969. attending the Scientific session of the Academy
Indian Academy of Medical Sciences and for presenting their papers. The papers
The Indian Academy of Medical Sciences should be submitted to the Executive Director
has been conducting postgraduate examinations not later than 15th September, 1969.
MINER’S DEATH CAUSED BY INH OVERDOSE
A FREAK accident, which resulted in the death of one African miner
and the poisoning of 199 others, occurred at the Doornfontein Gold Mine,
near Carletonville, early in November.
Instead of the usual purgative administered at their dressing stations
they received, in error, doses of isoniazid or INH, the well-known effective
drug used in TB treatment, which in therapeutic doses is harmless. They
developed stomach cramps and started vomiting and were taken to hospital
where one of them died later in the day. The others recovered.
The mine concerned is one of those which has in the past six years
been adding, with the consent of the mine workers, prophylactic INH to their
daily ration of marewu, a non-alcoholic maize drink popular among Africans.
This experimental scheme, reported in SANTA News (September, 1968) has
shown an 80 per cent decrease in TB incidence among the 41,000 African
miners who have taken part.
From SANTA NEWS, December, 1968
Ind. J. Tub., Vol. XVI, No. 2
UPHILL ANTI-T.B. FIGHT
A CERTAIN COMPLACENCY is discernible of late in the public atti-
tude to such dread scourges as tuberculosis. The development of powerful anti-
microbial drugs and the perfecting of effective and yet comparatively low- cost
domiciliary treatment for sufferers from this disease (as against the expen-
sive medical care in sanatoria) may have something to do with it. But the
figures of the high incidence of this killer (7 to 8 million cases) and deaths
(half a million) from it every year and the indequate facilities for its control
available, referred to by Dr. S. Chandrasekhar, Union Minister of State for
Health, at the 30th annual meeting of the Tuberculosis Association of India
should serve to warn against any relaxation in the war on this old and insidious
enemy. The plan of attack leans heavily on having at least one Control
Centre in each district from which radiate teams to diagnose and treat the
victims in the area. Only 171 of the 336 districts in India have such centres
and each of these control points can deal with only about 4,000 cases. The
inexorable arithmetic of it all is that as many as 5 to 6 million T.B sufferers go
without proper medical help. And coming from the Union Minister of Health
himself, this confession of inadequacy is indeed as alarming as it is authentic.
The obvious remedy is to quicken the tempo of enlargement of the clinics
and control centres. But the hurdles that have slowed down the process in
the past—lack of trained workers, funds, equipment and so on—are still there.
Fortunately, the voluntary T.B. Association of India has a wider network of
district branches, covering 207, than the official agency. Between these bran
ches and private doctors, the general practitioners practising in the area, much
may be done to fill the gap in more organised facilities. Dr. P.K. Duraiswami,
Chairman of the Association, has suggested the strengthening of the curriculum
of medical under-graduates, to include more expertise on the treatment of this
disease to help even G. Ps. to tackle T.B. patients with competence. Even as it
is, if every general practitioner refers suspected cases to the nearest clinic and
undertakes the supervision of the follow-up treatment and the education of
their contacts, they would be supplementing the organised network of
control centres. The Association which has been doing yeoman service to
publicise the perils of T.B. and harness public opinion to fight it may add more
sinews to its programme, if it can enlist both women’s organisations and the
private medical practitioners in its work to a much larger extent than now.
(Editorial in Hindu of April 21, 1969)
Ind. J. Tub., Vol. XVI, No. 2
The Indian Journal of Tuberculosis
Vol. XVI April 1969 Abst. No. 2
Sensitivity to Thiacetazone of Myco-Bacterium Of these 402 children, active tuberculosis
Tuberculosis Isolated in Algiers Practical was in 51 (12.5%) as shown radiologically and
J. Grosset; F. Rodriguez; M. Benhassine, P. The Heaf test was positive in only 11 of
Ghault and D. Larbaoui, Tubercle, Supplement, these children and the intradermal test using
Vol. 49, March, 1968. 100 tuberculin units was positive in a further
18 children. This confirms previous findings
There is, no corelation between ‘natural’ that tuberculin sensitivity is impaired in
sensititivity to thiacetazone and ‘natural’ malnourished children and suggest that a higher
sensitivity to ethionamide. However there is Co- dose of tuberculin is more likely to elicit a
relation between acquired resistance to ethio- positive response.
namide and resistance to thiacetazone. There
is however systematic cross resistance between H.B.D.
the two drugs.
Streptomycin Plus Thiacetazone (Thioacetazone)
H.B.D. Compared with Streptomycin Plus P.A.S.
and with Isoniazid Plus Thiacetazone in
Natural Sensitivity of M. Tuberculosis to the Treatment of Pulmonary Tuberculosis
Thiacetazone. in Rhodesia.
D.A. Mitchison, Tubercle, Supplement, Vol 49 L. Briggs et-al. R. W. Raddle, and Wallace Fox
March 1968. et-al. Tubercle, Land. (1968), 49, 48.
1 . Efficacy of thiacetazone plus isoniazid 220 patients were allocated at random to
varies from one area of the world to three treatment regimens.
2. Preliminary control trials should be
carried out when the use of this combi- Thiacetazone 150 mgm plus isoniazid 300
nation is envisaged on a major scale. mgm daily in a single tablet.
3. There is no association between the S.P.
results of pretreatment thiacetazone
sensitivity tests and the response to Streptomycin sulphate 1 gm intramuscu-
treatment. larly plus sodium PAS 15 Gm. daily in three
4. The routine use of such sensitivity test
is un-necessary when this combination S.T.
is widely used.
Streptomycin sulphate 1 gm. intra muscu-
H.B.D. larly plus thiacetazone 150 mgm daily in a
single tablet. All patients were treated in hos-
Tuberculin Test in Children with Malnutrition pital for six months. There were 171 (58 TH,
55 SP, 58 ST) patients who complied with
Aune V.C. Lloyd, Brit Med. Jur. 31st Aug. 1968. clinical criteria for admission and who had
strains of tubercle bacilli sensitive to isoniazid
In 402 children with severe malnutrition, and streptomycin. Ten patients died, five (two
Tuberculin test with Heaf’s multiple puncture SH, two SP, one ST) from active tuberculosis,
method as well as intradermal test with varying three (one SP, two ST) from non tuberculous
strength of old tuberculin were carried out causes but with active tuberculosis and two
Ind. J. Tub., Vol. XVI, No. 2
(both TH) from possible drug toxicity with Blood Levels of Isoniazid and of Its Methane
active tuberculosis. The sputum was negative Sulphonate Derivative in Rapid and Slow
on both cultures at six months in 72% of TH, Inactivators After Oral Administration.
82% of the SP and 33% of the ST patients. At
three months none (0%) of the 53 SP and 20 Aldo Baronti and Nella Manfredi, Tubercle,
(37%) of the 54 ST patients were excreting Lond., (1968), 49, 104.
streptomycin resistant organism (P<0.001). At
six months isoniazid resistant strains were Blood levels after oral administration of
obtained from 13% of 54 TH patients and 8% isoniazid and its methane sulphonate derivative
of 53 SP patients had culture with an RR of 4 (Methaniazid) were compared in a group of 20
or more of P.A.S. subjects consisting of 10 rapid and 10 slow
At six months 72% of 58 TH, 76% of 55
SP and 26% of 57 ST patients had favourable The two drugs were administered orally at a
response classified mainly on bacteriological dose of 5 mgm/Kgm calculated as isoniazid,
grounds. The difference between the ST and blood samples were drawn at three, six and
each of the other two regimens were statistically nine hours and assayed with the vertical diffu-
highly significant (P>0.0001). Cutaneous hyper sion method using Mycobactrium Tuberculosis
sensitivity occurred in 9% of the TH, of 10% H 37 R as test organism.
of the SP and 10% of the ST patients. It was
most severe in TH patients. Jaundice occurred The blood levels reached after administra-
in one patient in each of the three series and tion of methaniazid were significantly higher
dizziness was recorded only in the ST patients, than those obtained with isoniazid (PL < 0.001)
occurring in 11%. in slow inactivators the difference of blood
levels obtained with two drugs did not reach a
It is concluded that ST regimen was marke- significant level (0.10> P 70.05).
dly inferior. The SP and TH regimens were of H.B.D.
similar effectivness, the latter being an effective
oral regimen. Ethambutol Treatment of Tuberculosis in a
H.B.D. Controlled Trial
Aspergillus in Persistent Lung Cavities After Francis O. Segatra, Victor Lorian and David
Tuberculosis S. Sherman Scand. J. Resp. Dis : 493, 202.
A report from the Research Committee of The efficiency of ethambutol in combination
the British Tuberculosis Association, Tubercle, with INH and PAS in previously untreated
Land., (1968), 49, 1. cases has been evaluated in a controlled trial.
After excluding those who did not complete
Of 544 patients with persistent cavities of 3 months’ treatment, there were 47 patients
2.5 Cm. or more in diameter, 134 (25%) had in the ethambutol group and 20 in the control
a positive precipitin test. In 59 (11%) radio- group. Nearly 10% of the strains isolated
graphs showed typical appearances compatible from each group of the patients showed primary
with a aspergilloma and precipitin. A further resistance. The relief from clinical symptoms,
19 (4%) had less typical but highly suggestive radiological change and sputum conversion was
appearances and precipitines. In most of these marginally better in the control group. Further,
the precipitine test was strongly positive whilst the patients in the control group achieved
the result was weaker in those without such conversion in less time than the patients in the
radiographic evidence. The maximum preva- ethambutol group. One patient in the 20 in
lence of aspergillus infections occurred in those the control group developed acquired resistance
with cavitated tuberculosis of seven to 11 years to INH as compared with 9 of the 47 in the
duration and it became constant at a some- ethambutol group.
what lower level in those with longer standing
disease. Three patients in the control group deve-
loped severe intolerance to PAS and had to be
The affected patients had larger cavities and withdrawn from the study as against only one
cavities with thicker walls showed more pleural patient who developed jaundice in the 3rd
thickening and had more cough and were fre- week of treatment in the ethambutol group
quent recent haemoptysis. Five patients (1%) and had to be withdrawn. No visual distur-
without precipitins were found to have radio- bances were observed.
graphs typical of a mycetoma,
Ind. J, Tub., Vol. XVI, No, 2
Ethambutol in Initial Treatment The decrease in visual acuity was no more
in patients treated with ethambutol than in
Adil Sokmensuer, Transactions of the 27th others.
VA-Armed Forces, Pulmonary Diseases Research S.P.P.
Conference ; 1968, 3.
Ethambutol and Visual Acuity
Twenty three hospitals in United States co-
operated to determine the effectiveness as well Rae S. Newman, Transactions of the 27th VA-
as the toxicity of ethambutol as a substi- Armed Forces, Pulmonary Diseases Research
tute for PAS in initial treatment of Conference; 1968, 4.
pulmonary tuberculosis. Some of the patients
included in the study in the beginning The visual acuity in 2 groups of patients,
were given 6 mg per kg. of ethambutol but one given a dose of ethambutol 6 mg per kg.
subsequently the dose was increased to 15 mg and the other 16 mg per kg. body weight has
per kg. body weight. INH with low dose been compared in 1,219 patients. The decrease
ethambutol was as effective as INH and PAS in visual acuity was no more frequent in high
in reversing infectiousness for the first 12 weeks, dose group than in the low dose group.
but thereafter some patients in the former group
showed bacteriological reversions. In the case S.P.P.
of high dose ethambutol group, sputum con-
version throughout the trial period was equal The combined use of capreomycin and
to INH-PAS group and the bacilli in no case ethambutol in re-treatment of
developed resistance to INH. Adverse reactions pulmonary tuberculosis
severe enough to warrant withdrawal from the
drug schedule used, were seven times more Imasato Donomae: Amer. Rev. Resp. Dis.; 1968,
with PAS than with ethambutol. 98,699.
S.P.P. Capreomycin and ethambutol were given
for one year to 89 patients who underwent re-
Ethambutol in the Re-treatment of treatment for cavitary pulmonary tuberculosis
Pulmonary Tuberculosis and were excreting bacilli resistant to the
standard drugs. At 6 months, the cultures were
Hugh Kelly, Transactions of the 27th VA- negative in 75% whereas in the remaining 25% a
Armed Forces, Pulmonary Diseases Research gradual increase in the incidence of resistant
Conference ; 1968, 4. strains was seen with continuation of treatment.
Side effects of capreomycin viz difficulty in
Ninteen hospitals in United States partici- hearing, tinnitus and injection pain were re-
pated in a trial to test the effectiveness of ported by 7 patients though the symptoms were
ethambutal used in triple drug combination severe enough to require withdrawal of drug
for re-treatment of pulmonary tuberculosis. All in 3 cases only. The decline in visual acuity,
patients had received antimicrobial treatment abnormality in visual fields and eye strain
for at least 6 months prior to this trial. The occurred in 6 patients but ethambutol had to
drug regimens were used comprising of INH be withdrawn only in one case.
and various combinations of the second line
drugs. Nine of the ten regimens included Because of sclerotic nature of the lesion,
ethambutol also. Two hundred and seventy radiological regression of lesions was observed
seven patients completed the treatment for in 27.7% of the patients at 12 months. Slight
a stipulated period of 16 weeks. Eighty five to moderate decrease in the size of the cavity
percent of these patients were excreting bacilli was noticed in 36.3% of the patients. Closure
resistant to INH. The sputum conversion rate of cavities was obtained in 13.7% at 12 months
was 75%- The conversion rate showed treatment.
no significant difference in the various
regimens. Severe intolerance to ethambutol S.P.P.
warranting withdrawal of the drug appeared in
6 out of 203 patients receiving this drug. The antimicrobacterial activity of Rifampin
Capreomycin was given to 44 patients, and
none of these had any severe reaction. The Gladys L. Hobby, Tulita F. Lenert. Amer. Rev.
toxic reactions to the other second line drugs Resp. Dis., 1968, 97, 713.
used in the trial were fairly frequent. Ethiona-
mide had to be withdrawn in 13% and cyclo- A series of experiments were performed to
serine in 26%. compare the in vitro and in vivo activity of
Ind. J. Tub., Vol. XVI, No. 2
rifampin with that of INH. The data suggest chemotherapy. The total duration of treatment
that rifampin is approximately one half as was at least 18 months. Three drugs were
active as INH against INH-sensitive strains, used during the stay” in the hospital, one of
both in vitro and in vivo, but is active at least which was INH, streptomycin ethionamide or
in vitro against INH-resistant strains. pyrazinamide but the bacilli had to be fully
S.P.P. sensitive against all drugs used. During the
ambulatory treatment, only two drugs were
Re-treatment of patients with isoniazid-
resistant tuberculosis Sensitivity tests were available for 56
patients, out of which 48 were resistant to one
D.A. Fischer, William Lester, William E. Dye or two of the 3 standard drugs (viz INH
and Thomas S. Moulding. Amer. Rev. Resp. Dis.; streptomycin and PAS) and 8 were resistant to
1968, 97, 392. all three. Total cases resistant to INH were
45, to streptomycin 28 and to PAS 19. Of the
The results of treatment with various com- 45 strains resistant to INH, 15 were fully viru-
binations of second lines drugs and follow up lent and in the remaining 30, the virulence was
in 146 patients with INH resistant bacilli have attenuated.
been analysed. All patients were treated from
1960 to 1962, with a medium duration of 46 Out of the 69 chronic excretors, only 62
months follow up. After 120 days of treat- took drugs over 3 months. Of the remaining
ment, the sputum of 122 (83.5%) became 7, 4 died from a serious concurrent disease and
negative. The median time for conversion was 3 died from Tuberculosis within 3 months.
47th day of chemotherapy. There were 27 There were 25 more deaths up to the end of the
deaths up to January, 1966 and there were only study. Of the 37 patients surviving on 31st
7 survivors who failed to respond to treatment. March, 1967, 32 were converted and only 5
Thirty patients (20.5%) experienced bacterio- were still infectious. If deaths after 3 months
logical relapse during the period of observa- are included, the sputum conversion was
tion. Relapse occurred at a median time of obtained in 66%. Out of the 25 deaths
12 months after start of treatment. Significant amongst those who had been treated for more
toxic reactions were encountered in 42% of the than 3 months, 9 were converted at the time of
patients and appeared after a median duration death.
of more than 60 days of treatment with any
specific drug. Of the surviving patients whose There were 104 family contacts, of these 69
status was known in January 1966, 88% re- chronic excretors and they were also kept
mained consistently non-infectious. under serveillance. Two children, both BCG
vaccinated earlier, developed bacillary disease
S.P.P. during the period of observation. In one of
them the bacilli were resistant to INH and in
the second the bacilli were sensitive to all three
The Problem of the Chronic Excretor of drugs. None of the contacts over the age of
Tubercle Bacilli 15 developed tuberculosis. Although the role
of these chronic excretors in the dissemination
K. Styblo, A. Kubik, M. Langerova, E. Muthu of infection among general population cannot
Skova and K. Moravkova. Scand. J. Resp. Dis.; be estimated, the authors are of the opinion
1968, 49:3, 236. that they did not influence the incidence of
primary drug resistance in subsequent years
The management of chronic excretors of in the community to any appreciable extent.
tubercle bacilli in the district of Kolin, Czecho-
slovakia, with a population of 100,000 has S.P.P.
been studied. A person who had been excret-
ing tubercle bacilli persistently for a period of Primary Tuberculosis in Children
2 years or more was defined as ‘chronic ex-
cretor’. There were 53 such patients on 1st Morris Steiner, Raymond Zimmerman, Byung
October, 1960 when the study started and 16 Hak Park, Sudheer R. Shirall and Harry
more patients qualifying for the definition were Schmidt. Amer. Rev. Resp. Dis,; 1968, 98, 201.
added up to the end of the study on 31st
March, 1967. Of the 52 strains of M. Tuberculosis
isolated from children with primary disease, 3
All these patients were kept in the hospital strains were significantly resistant to INH and
for 9 to 15 months, followed by ambulatory 3 to streptomycin. The same prevalence of
Ind. J. Tub., Vol. XVI, No; 2
drug resistant strains was found in strains Effect of ultra violet irradiation on the acid
isolated from the corresponding source cases. fastness of drug-resistant mutants of tubercle
with special reference to the virulence of
All 137 strains of M. Tuberculosis isolated isoniazid-resistant strains
from all children, 9 were significantly resistant
to INH compared with 6 of the 79 strains Toy oho Murohashi and Konosuke Yoshida.
from the source cases. The prevalence of Amer. Rev. Resp. Dis.; 1968, 97, 283.
streptomycin and PAS resistant strains was
also similar. The study indicates that the level There is a high degree of co-relation bet-
of primary drug resistant infections among ween the degree of INH resistance, the viru-
children in the community under surveillance lence of the resistant bacilli and the effect of
is closely comparable to the level of drug ultra violet irradiation effect on acid-fastness
resistance among the adults” from whom the by a much shorter period of ultra violet
tuberculosis infection is acquired. irradiation than the virulent strains and their
streptomycin and PAS resistant mutants. The
S.P.P. factor that is suspected to be concerned appears
to be the cell wall structure of the bacillus
Spread of drug-resistant tubercle bacilli which becomes thinner in accordance with the
degree of resistance to INH.
E. Brander, K. Aho & J. Patiala. Amer. Rev. S.P.P.
Resp. Dis.; 1968, 98, 407.
Pseudocavities of the Lung
The source infection was studied in 50
adult cases of pulmonary tuberculosis caused Sanford E. Rabushka & Hiram T. Langs ton.
by primarily drug-resistant bacilli of human Amer. Rev. Resp. Dis.; 1968, 97, 644.
type. The patients were relatively young, with
a medium age of 25 years as compared to 41 The concept of pseudocavities in the lung
has been discussed. The high lipid content of
years in the case of all newly diagnosed cases. caseous material in the centre of round foci
Seven of the strains were resistant to all the often gives a shadow, indistinguishable radio-
three major drugs, 3 to two drugs and the logically from the shadow of a real cavity.
remaining 40 to one drug only. With the This concept has been proved through radiolo-
exception of two catalase—negative strains, the gical studies with ‘mock’ lesions attached to the
bacilli gave rise to progressive disease in the chest of volunteers before radiography.
Singing and the Dissemination of tuberculosis
A positive source of infection was found in
7 of the 11 patients more than 45 years old. Robert G. London. Rena Marie Roberts, Amar.
Only 3 of the 23 patients less than 25 years of Rev. Resp. Dis.; 1968, 98, 297.
age disclosed a possible source of infection
whereas 13 gave the history of old contact. The risk of droplet infection through singing
has been studied. Fewer droplets were ex-
The results indicated that the disease in the pelled during singing than during talking, but
a higher proportion of them were in the
majority of the young patients originated from smaller size range. The percentage of droplets
the old primary infection caused by drug still airborne as droplet nuclei after a 30,
resistant bacilli. The study also demonstrates minute settling period were 35.7, 6.4 and
usefulness of drug resistance as a microbial 48.9 for singing, talking and coughing res-
pectively. The very high proportion of smaller
marker in the epidemiological study of tuber- droplets expelled during singing would tend to
culosis. indicate high risk of infection through this
Ind. J. Tub., Vol. XVI, No. 2